Authorship：Corresponding author   Language：English   Publishing type：Research paper (scientific journal)   Publisher：Springer Nature Switzerland  

DOI： 10.1007/978-3-031-36190-6_4

researchmap

Authorship：Last author,　Corresponding author   Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/iiai-aai59060.2023.00074

researchmap

Authorship：Last author   Language：Japanese  

DOI： 10.11517/pjsai.JSAI2023.0_4Xin140

researchmap

Authorship：Last author   Language：Japanese  

DOI： 10.11517/pjsai.JSAI2023.0_3R1GS304

researchmap

Authorship：Lead author,　Corresponding author   Language：Japanese   Publishing type：Research paper (conference, symposium, etc.)  

researchmap

Authorship：Lead author,　Corresponding author   Language：English   Publishing type：Research paper (conference, symposium, etc.)  

researchmap

DOI： 10.1109/TCBB.2018.2814995

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Recent studies have shown that environmental DNA is found almost everywhere. Flower petal surfaces are an attractive tissue to use for investigation of the dispersal of environmental DNA in nature as they are isolated from the external environment until the bud opens and only then can the petal surface accumulate environmental DNA. Here, we performed a crowdsourced experiment, the "Ohanami Project", to obtain environmental DNA samples from petal surfaces of Cerasus × yedoensis 'Somei-yoshino' across the Japanese archipelago during spring 2015. C. × yedoensis is the most popular garden cherry species in Japan and clones of this cultivar bloom simultaneously every spring. Data collection spanned almost every prefecture and totaled 577 DNA samples from 149 collaborators. Preliminary amplicon-sequencing analysis showed the rapid attachment of environmental DNA onto the petal surfaces. Notably, we found DNA of other common plant species in samples obtained from a wide distribution; this DNA likely originated from the pollen of the Japanese cedar. Our analysis supports our belief that petal surfaces after blossoming are a promising target to reveal the dynamics of environmental DNA in nature. The success of our experiment also shows that crowdsourced environmental DNA analyses have considerable value in ecological studies.

DOI： 10.1007/s10265-018-1017-x

PubMed

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：BioMed Central Ltd.  

Background: Comprehensively understanding the dynamics of biological systems is among the biggest current challenges in biology and medicine. To acquire this understanding, researchers have measured the time-series expression profiles of cell lines of various organisms. Biological technologies have also drastically improved, providing a huge amount of information with support from bioinformatics and systems biology. However, the transitions between the activation and inactivation of gene regulations, at the temporal resolution of single time points, are difficult to extract from time-course gene expression profiles. Results: Our proposed method reports the activation period of each gene regulation from gene expression profiles and a gene regulatory network. The correctness and effectiveness of the method were validated by analyzing the diauxic shift from glucose to lactose in Escherichia coli. The method completely detected the three periods of the shift
1) consumption of glucose as nutrient source, 2) the period of seeking another nutrient source and 3) consumption of lactose as nutrient source. We then applied the method to mouse adipocyte differentiation data. Cell differentiation into adipocytes is known to involve two waves of the gene regulation cascade, and sub-waves are predicted. From the gene expression profiles of the cell differentiation process from ES to adipose cells (62 time points), our method acquired four periods
three periods covering the two known waves of the cascade, and a final period of gene regulations when the differentiation to adipocytes was completed. Conclusions: Our proposed method identifies the transitions of gene regulations from time-series gene expression profiles. Dynamic analyses are essential for deep understanding of biological systems and for identifying the causes of the onset of diseases such as diabetes and osteoporosis. The proposed method can greatly contribute to the progress of biology and medicine.

DOI： 10.1186/s12859-018-2072-y

Scopus

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Background: Epigenetic factors associated with the development of autoimmune diseases are unclear. Monozygotic twin pairs discordant for positive antithyroglobulin autoantibodies (TgAb) are useful to examine the epigenetic factors because of their identical genetic background. This study aimed to clarify the discordant epigenetic differences affecting the development of TgAb. Methods: Subjects were selected from 257 Japanese monozygotic twins, recruited from the registry established by the Center for Twin Research at Osaka University. TgAb positive concordant (PC) pairs were 5.7% (four pairs) and 9.6% (18 pairs) of male and female pairs, respectively. TgAb discordant (DC) pairs were 11.4% (eight pairs) and 8.0% (15 pairs) of male and female pairs, respectively. TgAb negative concordant (NC) pairs were 78.6% (55 pairs) of male pairs and 74.3% (139 pairs) of female pairs. To perform stricter grouping, the cut-off value for positive TgAb was set to 50.0 IU/mL (TgAb negative: <28.0 IU/mL; TgAb positive: ≥50.0 IU/mL; TgAb borderline: ≥28.0 IU/mL and <50.0 IU/mL). Nineteen discordant (6 male and 13 female pairs) and 185 concordant pairs (48 male and 137 female pairs) for TgAb positivity were finally examined. DNA methylation levels of genomic DNA were evaluated using the Infinium HumanMethylation450 BeadChip Kit. Gene polymorphisms were also genotyped using the Omni5-4 BeadChip Kit to clarify genetic background specific for discordant twins. Results: No CpG sites were found with significant within-pair differences of methylation levels in TgAb DC pairs after correction for multiple comparisons. However, 155 polymorphisms specific for TgAb DC pairs were significantly different in genotype frequencies from those of concordant pairs, and none of them were located on the HLA region of chromosome 6. In TgAb DC pairs with some specific genotypes of these polymorphisms, four CpG sites were observed exhibiting significant within-pair differences in each DC pair, even after correction for multiple comparisons. Conclusions: This study found that the genetic background specific for TgAb DC twins who are susceptible to epigenetic changes are different from that specific for TgAb PC twins, and it clarified the genotype-based epigenetic differences in TgAb DC monozygotic twins.

DOI： 10.1089/thy.2017.0273

PubMed

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：AMER MEDICAL ASSOC  

IMPORTANCE Application of deep learning algorithms to whole-slide pathology images can potentially improve diagnostic accuracy and efficiency.
OBJECTIVE Assess the performance of automated deep learning algorithms at detecting metastases in hematoxylin and eosin-stained tissue sections of lymph nodes of women with breast cancer and compare it with pathologists' diagnoses in a diagnostic setting.
DESIGN, SETTING, AND PARTICIPANTS Researcher challenge competition (CAMELYON16) to develop automated solutions for detecting lymph node metastases (November 2015-November 2016). A training data set of whole-slide images from 2 centers in the Netherlands with (n = 110) and without (n = 160) nodal metastases verified by immunohistochemical staining were provided to challenge participants to build algorithms. Algorithm performance was evaluated in an independent test set of 129 whole-slide images (49 with and 80 without metastases). The same test set of corresponding glass slides was also evaluated by a panel of 11 pathologists with time constraint (WTC) from the Netherlands to ascertain likelihood of nodal metastases for each slide in a flexible 2-hour session, simulating routine pathology workflow, and by 1 pathologist without time constraint (WOTC).
EXPOSURES Deep learning algorithms submitted as part of a challenge competition or pathologist interpretation.
MAIN OUTCOMES AND MEASURES The presence of specific metastatic foci and the absence vs presence of lymph node metastasis in a slide or image using receiver operating characteristic curve analysis. The 11 pathologists participating in the simulation exercise rated their diagnostic confidence as definitely normal, probably normal, equivocal, probably tumor, or definitely tumor.
RESULTS The area under the receiver operating characteristic curve (AUC) for the algorithms ranged from 0.556 to 0.994. The top-performing algorithm achieved a lesion-level, true-positive fraction comparable with that of the pathologist WOTC (72.4%[95% CI, 64.3%-80.4%]) at a mean of 0.0125 false-positives per normal whole-slide image. For the whole-slide image classification task, the best algorithm (AUC, 0.994 [95% CI, 0.983-0.999]) performed significantly better than the pathologists WTC in a diagnostic simulation (mean AUC, 0.810 [range, 0.738-0.884]; P&lt;.001). The top 5 algorithms had a mean AUC that was comparable with the pathologist interpreting the slides in the absence of time constraints (mean AUC, 0.960 [range, 0.923-0.994] for the top 5 algorithms vs 0.966 [95% CI, 0.927-0.998] for the pathologist WOTC).
CONCLUSIONS AND RELEVANCE In the setting of a challenge competition, some deep learning algorithms achieved better diagnostic performance than a panel of 11 pathologists participating in a simulation exercise designed to mimic routine pathology workflow; algorithm performance was comparable with an expert pathologist interpreting whole-slide images without time constraints. Whether this approach has clinical utility will require evaluation in a clinical setting.

DOI： 10.1001/jama.2017.14585

Web of Science

researchmap

Language：English   Publisher：IEEE Computer Society  

DOI： 10.1109/BIBE.2016.51

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：IMPERIAL COLLEGE PRESS  

Comprehensively understanding the dynamics of biological systems is one of the greatest challenges in biology. Vastly improved biological technologies have provided vast amounts of information that must be understood by bioinformatics and systems biology researchers. Gene regulations have been frequently modeled by ordinary differential equations or graphical models based on time-course gene expression profiles. The state-of-the-art computational approaches for analyzing gene regulations assume that their models are same throughout time-course experiments. However, these approaches cannot easily analyze transient changes at a time point, such as diauxic shift. We propose a score that analyzes the gene regulations at each time point. The score is based on the information gains of information criterion values. The method detects the shifts in gene regulatory networks (GRNs) during time-course experiments with single-time-point resolution. The effectiveness of the method is evaluated on the diauxic shift from glucose to lactose in Escherichia coli. Gene regulation shifts were detected at two time points: the first corresponding to the time at which the growth of E. coli ceased and the second corresponding to the end of the experiment, when the nutrient sources (glucose and lactose) had become exhausted. According to these results, the proposed score and method can appropriately detect the time of gene regulation shifts. The method based on the proposed score provides a new tool for analyzing dynamic biological systems. Because the score value indicates the strength of gene regulation at each time point in a gene expression profile, it can potentially infer hidden GRNs from time-course experiments.

DOI： 10.1142/S0219720015430027

Web of Science

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

A similarity measure between statute books of local governments that can help reveal suggestive similarities is proposed. The regulations of a local government are stored in a statute book, and they are categorized in a layered structure. The layered structure can be described as an ordered tree in computer science, and we define the similarity of statute books as the tree edit distance between two trees. We have calculated the similarities among statute books of the 47 Japanese prefectures and plotted them on a plane using multi-dimensional scaling. The results visually indicate the relationships of similarities among them, and there are several outlier prefectures and clusters. This will help find local governments with similar regulations, which will facilitate the writing or revision of statutes, especially in small local governments, which are typically short staffed.

DOI： 10.1109/KSE.2015.57

Web of Science

researchmap

Language：English  

researchmap

Language：English  

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：PEERJ INC  

A vast amount of metagenomic data has been obtained by extracting multiple genomes simultaneously from microbial communities, including genomes from uncultivable microbes. By analyzing these metagenomic data, novel microbes are discovered and new microbial functions are elucidated. The first step in analyzing these data is sequenced-read classification into reference genomes from which each read can be derived. The Naive Bayes Classifier is a method for this classification. To identify the derivation of the reads, this method calculates a score based on the occurrence of a DNA sequence motif in each reference genome. However, large differences in the sizes of the reference genomes can bias the scoring of the reads. This bias might cause erroneous classification and decrease the classification accuracy. To address this issue, we have updated the Naive Bayes Classifier method usingmultiple sets of occurrence profiles for each reference genome by normalizing the genome sizes, dividing each genome sequence into a set of subsequences of similar length and generating profiles for each subsequence. This multiple profile strategy improves the accuracy of the results generated by the Naive Bayes Classifier method for simulated and Sargasso Sea datasets.

DOI： 10.7717/peerj.559

Web of Science

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

近年、バイオイメージング技術の進歩によって，様々な条件下での細胞や組織を高並列かつ長時間の動画として観察することが出来るようになっている．また様々なタンパク質の発現にそれぞれ異なる蛍光色素をプローブとしてカップリングする技術も進展しており，創薬研究における薬剤候補の選別や生命科学における表現型解析などへ応用されている．大量の動画から，様々な特徴量を抽出し細胞の特性の変化を定量的に解析するためには、自動的に精度良く細胞の追跡を行う必要がある．本研究では，複数種類の蛍光標識により多重染色された細胞をマルチチャンネルの蛍光顕微鏡で経時観察して得られる動画を対象として，パーティクルフィルタを用いた細胞の追跡手法を提案し，その有効性を評価するとともに得られた結果の考察を行う．

CiNii Books

researchmap

Language：Japanese  

CiNii Books

researchmap

Language：Japanese  

CiNii Books

researchmap

Language：English  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Background: Identifying the differences between gene regulatory networks under varying biological conditions or external stimuli is an important challenge in systems biology. Several methods have been developed to reverse-engineer a cellular system, called a gene regulatory network, from gene expression profiles in order to understand transcriptomic behavior under various conditions of interest. Conventional methods infer the gene regulatory network independently from each of the multiple gene expression profiles under varying conditions to find the important regulatory relations for understanding cellular behavior. However, the inferred networks with conventional methods include a large number of misleading relations, and the accuracy of the inference is low. This is because conventional methods do not consider other related conditions, and the results of conventional methods include considerable noise due to the limited number of observation points in each expression profile of interest. Results: We propose a more accurate method for estimating key gene regulatory networks for understanding cellular behavior under various conditions. Our method utilizes multiple gene expression profiles that compose a tree structure under varying conditions. The root represents the original cellular state, and the leaves represent the changed cellular states under various conditions. By using this tree-structured gene expression profiles, our method more powerfully estimates the networks that are key to understanding the cellular behavior of interest under varying conditions. Conclusion: We confirmed that the proposed method in cell differentiation was more rigorous than the conventional method. The results show that our assumptions as to which relations are unimportant for understanding the differences of cellular states in cell differentiation are appropriate, and that our method can infer more accurately the core networks of the cell types. © 2012 Elsevier B.V.

DOI： 10.1016/j.gene.2012.11.090

Scopus

PubMed

researchmap

Language：English   Publisher：Japanese Society for Medical and Biological Engineering  

DOI： 10.11239/jsmbe.51.R-166

researchmap

Language：English  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：English  

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：BIOMED CENTRAL LTD  

Background: Bayesian networks (BNs) have been widely used to estimate gene regulatory networks. Many BN methods have been developed to estimate networks from microarray data. However, two serious problems reduce the effectiveness of current BN methods. The first problem is that BN-based methods require huge computational time to estimate large-scale networks. The second is that the estimated network cannot have cyclic structures, even if the actual network has such structures.
Results: In this paper, we present a novel BN-based deterministic method with reduced computational time that allows cyclic structures. Our approach generates all the combinational triplets of genes, estimates networks of the triplets by BN, and unites the networks into a single network containing all genes. This method decreases the search space of predicting gene regulatory networks without degrading the solution accuracy compared with the greedy hill climbing (GHC) method. The order of computational time is the cube of number of genes. In addition, the network estimated by our method can include cyclic structures.
Conclusions: We verified the effectiveness of the proposed method for all known gene regulatory networks and their expression profiles. The results demonstrate that this approach can predict regulatory networks with reduced computational time without degrading the solution accuracy compared with the GHC method.

DOI： 10.1186/1471-2164-13-S1-S12

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Alternative splicing plays an important role in eukaryotic gene expression by producing diverse proteins from a single gene. Predicting how genes are transcribed is of great biological interest. To this end, massively parallel whole transcriptome sequencing, often referred to as RNA-Seq, is becoming widely used and is revolutionizing the cataloging isoforms using a vast number of short mRNA fragments called reads. Conventional RNA-Seq analysis methods typically align reads onto a reference genome (mapping) in order to capture the form of isoforms that each gene yields and how much of every isoform is expressed from an RNA-Seq dataset. However, a considerable number of reads cannot be mapped uniquely. Those so-called multireads that are mapped onto multiple locations due to short read length and analogous sequences inflate the uncertainty as to how genes are transcribed. This causes inaccurate gene expression estimations and leads to incorrect isoform prediction. To cope with this problem, we propose a method for isoform prediction by iterative mapping. The positions from which multireads originate can be estimated based on the information of expression levels, whereas quantification of isoform-level expression requires accurate mapping. These procedures are mutually dependent, and therefore remapping reads is essential. By iterating this cycle, our method estimates gene expression levels more precisely and hence improves predictions of alternative splicing. Our method simultaneously estimates isoform-level expressions by computing how many reads originate from each candidate isoform using an EM algorithm within a gene. To validate the effectiveness of the proposed method, we compared its performance with conventional methods using an RNA-Seq dataset derived from a human brain. The proposed method had a precision of 66.7% and outperformed conventional methods in terms of the isoform detection rate.

DOI： 10.2197/ipsjtbio.5.27

Scopus

researchmap

Language：English   Publisher：Information and Media Technologies Editorial Board  

Alternative splicing plays an important role in eukaryotic gene expression by producing diverse proteins from a single gene. Predicting how genes are transcribed is of great biological interest. To this end, massively parallel whole transcriptome sequencing, often referred to as RNA-Seq, is becoming widely used and is revolutionizing the cataloging isoforms using a vast number of short mRNA fragments called reads. Conventional RNA-Seq analysis methods typically align reads onto a reference genome (mapping) in order to capture the form of isoforms that each gene yields and how much of every isoform is expressed from an RNA-Seq dataset. However, a considerable number of reads cannot be mapped uniquely. Those so-called multireads that are mapped onto multiple locations due to short read length and analogous sequences inflate the uncertainty as to how genes are transcribed. This causes inaccurate gene expression estimations and leads to incorrect isoform prediction. To cope with this problem, we propose a method for isoform prediction by iterative mapping. The positions from which multireads originate can be estimated based on the information of expression levels, whereas quantification of isoform-level expression requires accurate mapping. These procedures are mutually dependent, and therefore remapping reads is essential. By iterating this cycle, our method estimates gene expression levels more precisely and hence improves predictions of alternative splicing. Our method simultaneously estimates isoform-level expressions by computing how many reads originate from each candidate isoform using an EM algorithm within a gene. To validate the effectiveness of the proposed method, we compared its performance with conventional methods using an RNA-Seq dataset derived from a human brain. The proposed method had a precision of 66.7% and outperformed conventional methods in terms of the isoform detection rate.

DOI： 10.11185/imt.7.921

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Imperial College Press  

Background: Bayesian networks (BNs) have been widely used to estimate gene regulatory networks. Many BN methods have been developed to estimate networks from microarray data. However, two serious problems reduce the effectiveness of current BN methods. The first problem is that BN-based methods require huge computational time to estimate large-scale networks. The second is that the estimated network cannot have cyclic structures, even if the actual network has such structures.
Results: In this paper, we present a novel BN-based deterministic method with reduced computational time that allows cyclic structures. Our approach generates all the combinational triplets of genes, estimates networks of the triplets by BN, and unites the networks into a single network containing all genes. This method decreases the search space of predicting gene regulatory networks without degrading the solution accuracy compared with the greedy hill climbing (GHC) method. The order of computational time is the cube of number of genes. In addition, the network estimated by our method can include cyclic structures.
Conclusions: We verified the effectiveness of the proposed method for all known gene regulatory networks and their expression profiles. The results demonstrate that this approach can predict regulatory networks with reduced computational time without degrading the solution accuracy compared with the GHC method.

DOI： 10.1186/1471-2164-13-S1-S12

Scopus

PubMed

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：IMPERIAL COLLEGE PRESS  

The S-system model is one of the nonlinear differential equation models of gene regulatory networks, and it can describe various dynamics of the relationships among genes. If we successfully infer rigorous S-system model parameters that describe a target gene regulatory network, we can simulate gene expressions mathematically. However, the problem of finding an optimal S-system model parameter is too complex to be solved analytically. Thus, some heuristic search methods that offer approximate solutions are needed for reducing the computational time. In previous studies, several heuristic search methods such as Genetic Algorithms (GAs) have been applied to the parameter search of the S-system model. However, they have not achieved enough estimation accuracy. One of the conceivable reasons is that the mechanisms to escape local optima. We applied an Immune Algorithm (IA) to search for the S-system parameters. IA is also a heuristic search method, which is inspired by the biological mechanism of acquired immunity. Compared to GA, IA is able to search large solution space, thereby avoiding local optima, and have multiple candidates of the solutions. These features work well for searching the S-system model. Actually, our algorithm showed higher performance than GA for both simulation and real data analyses.

DOI： 10.1142/S0219720011005768

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：BIOMED CENTRAL LTD  

Background: With the advent of next-generation sequencers, the growing demands to map short DNA sequences to a genome have promoted the development of fast algorithms and tools. The tools commonly used today are based on either a hash table or the suffix array/Burrow-Wheeler transform. These algorithms are the best suited to finding the genome position of exactly matching short reads. However, they have limited capacity to handle the mismatches. To find n-mismatches, they requires O(2(n)) times the computation time of exact matches. Therefore, acceleration techniques are required.
Results: We propose a hash-based method for genome mapping that reduces the number of hash references for finding mismatches without increasing the size of the hash table. The method regards DNA subsequences as words on Galois extension field GF(2(2)) and each word is encoded to a code word of a perfect Hamming code. The perfect Hamming code defines equivalence classes of DNA subsequences. Each equivalence class includes subsequence whose corresponding words on GF(2(2)) are encoded to a corresponding code word. The code word is used as a hash key to store these subsequences in a hash table. Specifically, it reduces by about 70% the number of hash keys necessary for searching the genome positions of all 2-mismatches of 21-base-long DNA subsequence.
Conclusions: The paper shows perfect hamming code can reduce the number of hash references for hash-based genome mapping. As the computation time to calculate code words is far shorter than a hash reference, our method is effective to reduce the computation time to map short DNA sequences to genome. The amount of data that DNA sequencers generate continues to increase and more accurate genome mappings are required. Thus our method will be a key technology to develop faster genome mapping software.

DOI： 10.1186/1471-2164-12-S3-S8

Web of Science

researchmap

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

CiNii Books

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：SPRINGER  

Peritoneal relapse is the most common pattern of tumor progression in advanced gastric cancer. Clinicopathological findings are sometimes inadequate for predicting peritoneal relapse. The aim of this study was to identify patients at high risk of peritoneal relapse in a prospective study based on molecular prediction.
RNA samples from 141 primary gastric cancer tissues after curative surgery were profiled using oligonucleotide microarrays covering 30,000 human probes. Firstly, we constructed a molecular prediction system and validated its robustness and prognostic validity by 500 times multiple validation by repeated random sampling in a retrospective set of 56 (38 relapse-free and 18 peritoneal-relapse) patients. Secondly, we applied this prediction to 85 patients of the prospective set to assess predictive accuracy and prognostic validity.
In the retrospective phase, repeated random validation yielded similar to 68% predictive accuracy and a 22-gene expression profile associated with peritoneal relapse was identified. The prediction system identified patients with poor prognosis. In the prospective phase, the molecular prediction yielded 76.9% overall accuracy. Kaplan-Meier analysis of peritoneal-relapse-free survival showed a significant difference between the "good signature group" and "poor signature group" (log-rank p = 0.0017). Multivariate analysis by Cox regression hazards model identified the molecular prediction as the only independent prognostic factor for peritoneal relapse.
Gene expression profile inherent to primary gastric cancer tissues can be useful in prospective prediction of peritoneal relapse after curative surgery, potentially allowing individualized postoperative management to improve the prognosis of patients with advanced gastric cancer.

DOI： 10.1245/s10434-009-0854-1

Web of Science

researchmap

Language：English   Publisher：Information and Media Technologies Editorial Board  

To identify protein-protein interaction pairs with high accuracy, we propose a method for predicting these interactions based on characteristics obtained from protein-protein docking evaluations. Previous studies assumed that the required protein affinity strength for an interaction was not dependent on protein functions. However, the protein affinity strength appears to differ with different docking schemes, such as rigid-body or flexible docking, and these schemes may be related to protein functions. Thus, we propose a new scoring system that is based on statistical analysis of affinity score distributions sampled by their protein functions. As a result, of all possible protein pair combinations, a newly developed method improved prediction accuracy of F-measures. In particular, for bound antibody-antigen pairs, we obtained 50.0% recall (=sensitivity) with higher F-measures compared with previous studies. In addition, by combining two proposed scoring systems, Receptor-Focused Z-scoring and Ligand-Focused Z-scoring, further improvement was achieved. This result suggested that the proposed prediction method improved the prediction accuracy (i.e., F-measure), with few false positives, by taking biological functions of protein pairs into consideration.

DOI： 10.11185/imt.5.489

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

PubMed

researchmap

researchmap

Language：Japanese   Publishing type：Research paper (other academic)  

researchmap

Language：English   Publishing type：Research paper (other academic)  

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

Language：Japanese   Publishing type：Research paper (other academic)   Publisher：Information Processing Society of Japan (IPSJ)  

The amount of compounds in public databases goes on increasing. A structure key and Tanimoto coefficient are often used for similarity searches in compound databases. As the number of compounds increases, the search time increases. In this research, we propose a method for refinement of the calculation range by divided structure key, and show the effectiveness by measuring the number of calculation with the data of database.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00058866/

Language：English   Publisher：Information Processing Society of Japan  

The number of biological databases has been increasing rapidly as a result of progress in biotechnology. As the amount and heterogeneity of biological data increase, it becomes more difficult to manage the data in a few centralized databases. Moreover, the number of sites storing these databases is getting larger, and the geographic distribution of these databases has become wider. In addition, biological research tends to require a large amount of computational resources, i.e., a large number of computing nodes. As such, the computational demand has been increasing with the rapid progress of biological research. Thus, the development of methods that enable computing nodes to use such widely-distributed database sites effectively is desired. In this paper, we propose a method for providing data from the database sites to computing nodes. Since it is difficult to decide which program runs on a node and which data are requested as their inputs in advance, we have introduced the notion of "data-staging" in the proposed method. Data-staging dynamically searches for the input data from the database sites and transfers the input data to the node where the program runs. We have developed a prototype system with data-staging using grid middleware. The effectiveness of the prototype system is demonstrated by measurement of the execution time of similarity search of several-hundred gene sequences against 527 prokaryotic genome data.

DOI： 10.2197/ipsjdc.4.250

researchmap

Language：English   Publisher：Information Processing Society of Japan  

Chemical and biological activities of compounds provide valuable information for discovering new drugs. The compound fingerprint that is represented by structural information of the activities is used for candidates for investigating similarity. However, there are several problems with predicting accuracy from the requirement in the compound structural similarity. Although the amount of compound data is growing rapidly, the number of well-annotated compounds, e.g., those in the MDL Drug Data Report (MDDR)database, has not increased quickly. Since the compounds that are known to have some activities of a biological class of the target are rare in the drug discovery process, the accuracy of the prediction should be increased as the activity decreases or the false positive rate should be maintained in databases that have a large number of un-annotated compounds and a small number of annotated compounds of the biological activity. In this paper, we propose a new similarity scoring method composed of a combination of the Tanimoto coefficient and the proximity measure of random forest. The score contains two properties that are derived from unsupervised and supervised methods of partial dependence for compounds. Thus, the proposed method is expected to indicate compounds that have accurate activities. By evaluating the performance of the prediction compared with the two scores of the Tanimoto coefficient and the proximity measure, we demonstrate that the prediction result of the proposed scoring method is better than those of the two methods by using the Linear Discriminant Analysis (LDA) method. We estimate the prediction accuracy of compound datasets extracted from MDDR using the proposed method. It is also shown that the proposed method can identify active compounds in datasets including several un-annotated compounds.

DOI： 10.2197/ipsjdc.4.238

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Japanese Society for Bioinformatics  

We present a new method to describe tissue-specific function that leverages the advantage of the Cap Analysis of Gene Expression (CAGE) data. The CAGE expression data represent the number of mRNAs of each gene in a sample. The feature enables us to compare or add the expression amount of genes in the sample. As usual methods compared the gene expression values among tissues for each gene respectively and ruled out to compare them among genes, they have not exploited the feature to reveal tissue specifi city. To utilize the feature, we used Gene Ontology terms (GO-terms) as unit to sum up the expression values and described specificities of tissues by them. We regard GO-terms as events that occur in the tissue according to probabilities that are defined by means of the CAGE. Our method is applied to mouse CAGE data on 22 tissues. Among them, we show the results of molecular functions and cellular components on liver. We also show the most expressed genes in liver to compare with our method. The results agree well with well-known specific functions such as amino acid metabolisms of liver. Moreover, the difference of inter-cellular junction among liver, lung, heart, muscle and prostate gland are apparently observed. The results of our method provide researchers a clue to the further research of the tissue roles and the deeper functions of the tissue-specific genes. All the results and supplementary materials are available via our web site.

DOI： 10.11234/gi1990.19.154

PubMed

researchmap

Language：English  

researchmap

researchmap

Language：Japanese  

researchmap

Language：English   Publisher：Information Processing Society of Japan  

In bioinformatics, dealing with tools to analyze biological data becomes important. Those tools are provided by various institutions and the number of the tools is rapidly increasing. Recently many institutions have been offering those tools and access to databases with Web service technologies. The workflow technology is one of the ways to manage those tools and it is becoming available in bioinformatics. In order to compose workflows, several research groups develop and provide workflow composing tools. And consequently, the concept "Workflow Reuse" is also arisen in order to help workflow composition. Nevertheless it is still difficult to search for the reusable workflows from the repository of workflows in the current situation. In this paper, we propose a method to extract reusable workflows from the repository by using currently available information. We could extract some functionally similar workflows as reusable ones. By extracting reusable workflows efficiently, researchers can compose their workflow more easily.

DOI： 10.2197/ipsjdc.3.164

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

The accumulation of gene-expression profiles can allow an inference of a gene regulatory network by using a profile measured under a number of experimental conditions. However, in case of applying to such profile, the conventional methods such as module network model may not perform an inference accurately enough, because of following two facts. 1) Module network can accurately perform an inference only for regulated genes that show similar gene-expression patterns under almost all experimental conditions. 2) In a gene-expression profile that includes more conditions, fewer genes show similar gene-expression patterns. To alleviate the accuracy loss, we utilized a biclustering result for an inference. We performed an inference for regulated genes that were included in a detected bicluster by using gene-expression patterns only under experimental conditions included in the bicluster. We demonstrate the effectiveness of our method by applying to inferences of gene regulatory networks by using various gene-expression profiles of budding yeast.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00017157/

Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：PUBLIC LIBRARY SCIENCE  

T he international FANTOM consortium aims to produce a comprehensive picture of the mammalian transcriptome, based upon an extensive cDNA collection and functional annotation of full-length enriched cDNAs. The previous dataset, FANTOM(2), comprised 60,770 full- length enriched cDNAs. Functional annotation revealed that this cDNA dataset contained only about half of the estimated number of mouse protein- coding genes, indicating that a number of cDNAs still remained to be collected and identified. To pursue the complete gene catalog that covers all predicted mouse genes, cloning and sequencing of full- length enriched cDNAs has been continued since FANTOM2. In FANTOM3, 42,031 newly isolated cDNAs were subjected to functional annotation, and the annotation of 4,347 FANTOM2 cDNAs was updated. To accomplish accurate functional annotation, we improved our automated annotation pipeline by introducing new coding sequence prediction programs and developed a Web- based annotation interface for simplifying the annotation procedures to reduce manual annotation errors. Automated coding sequence and function prediction was followed with manual curation and review by expert curators. A total of 102,801 full- length enriched mouse cDNAs were annotated. Out of 102,801 transcripts, 56,722 were functionally annotated as protein coding ( including partial or truncated transcripts), providing to our knowledge the greatest current coverage of the mouse proteome by full- length cDNAs. The total number of distinct non- protein- coding transcripts increased to 34,030. The FANTOM3 annotation system, consisting of automated computational prediction, manual curation, and. nal expert curation, facilitated the comprehensive characterization of the mouse transcriptome, and could be applied to the transcriptomes of other species.

DOI： 10.1371/journal.pgen.0020062

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：PUBLIC LIBRARY SCIENCE  

With the advancement of genome research, it is becoming clear that genes are not distributed on the genome in random order. Clusters of genes distributed at localized genome positions have been reported in several eukaryotes. Various correlations have been observed between the expressions of genes in adjacent or nearby positions along the chromosomes depending on tissue type and developmental stage. Moreover, in several cases, their transcripts, which control epigenetic transcription via processes such as transcriptional interference and genomic imprinting, occur in clusters. It is reasonable that genomic regions that have similar mechanisms show similar expression patterns and that the characteristics of expression in the same genomic regions differ depending on tissue type and developmental stage. In this study, we analyzed gene expression patterns using the cap analysis gene expression ( CAGE) method for exploring systematic views of the mouse transcriptome. Counting the number of mapped CAGE tags for fixed-length regions allowed us to determine genomic expression levels. These expression levels were normalized, quantified, and converted into four types of descriptors, allowing the expression patterns along the genome to be represented by character strings. We analyzed them using dynamic programming in the same manner as for sequence analysis. We have developed a novel algorithm that provides a novel view of the genome from the perspective of genomic positional expression. In a similarity search of expression patterns across chromosomes and tissues, we found regions that had clusters of genes that showed expression patterns similar to each other depending on tissue type. Our results suggest the possibility that the regions that have sense - antisense transcription show similar expression patterns between forward and reverse strands.

DOI： 10.1371/journal.pgen.0020044

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：WORLD SCIENTIFIC PUBL CO PTE LTD  

With the rapid progress of the human genome project and related analyses, a huge amount of sequence data has been generated and a substantial number of methods has been proposed for predicting the potential functions based on sequence homology, functional patterns (motifs), domain information and so forth. It is often the case that actual processes of these biological analyses are not straightforward but rather complicated. In order to solve this problem, we propose a framework to virtualize and integrate various biological resources such as programs, databases and experimental data on the Grid environment. We show how our architecture makes it possible to improve the complicated process of biological analyses.

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：AMER ASSOC ADVANCEMENT SCIENCE  

This study describes comprehensive polling of transcription start and termination sites and analysis of previously unidentified full-length complementary DNAs derived from the mouse genome. We identify the 5' and 3' boundaries of 181,047 transcripts with extensive variation in transcripts arising from alternative promoter usage, splicing, and polyadenylation. There are 16,247 new mouse protein-coding transcripts, including 5154 encoding previously unidentified proteins. Genomic mapping of the transcriptome reveals transcriptional forests, with overlapping transcription on both strands, separated by deserts in which few transcripts are observed. The data provide a comprehensive platform for the comparative analysis of mammalian transcriptional regulation in differentiation and development.

DOI： 10.1126/science.1112014

Web of Science

researchmap

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：OXFORD UNIV PRESS  

Motivation: Clustering of protein sequences is widely used for the functional characterization of proteins. However, it is still not easy to cluster distantly-related proteins, which have only regional similarity among their sequences. It is therefore necessary to develop an algorithm for clustering such distantly-related proteins.
Results: We have developed a time and space efficient clustering algorithm. It uses a graph representation where its vertices and edges denote proteins and their sequence similarities above a certain cutoff score, respectively. It repeatedly partitions the graph by removing edges that have small weights, which correspond to low sequence similarities. To find the appropriate partitions, we introduce a score combining the normalized cut and a locally minimal cut capacities. Our method is applied to the entire 40 703 human proteins in SWISS-PROT and TrEMBL. The resulting clusters shows a 76% recall (20 529 proteins) of the 26 917 classified by InterPro. It also finds relationships not found by other clustering methods.

DOI： 10.1093/bioinformatics/btg397

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：2  

Recently, gene expression data under various conditions have largely been obtained by the utilization of the DNA microarrays and oligonucleotide arrays. There have been emerging demands to analyze the function of genes from the gene expression profiles. For clustering genes from their expression profiles, hierarchical clustering has been widely used. The clustering method represents the relationships of genes as a tree structure by connecting genes using their similarity scores based on the Pearson correlation coefficient. But the clustering method is sensitive to experimental noise.<BR>To cope with the problem, we propose another type of clustering method (the <I>p</I>-quasi complete linkage clustering). We apply this method to the gene expression data of yeast cell-cycles and human lung cancer. The effectiveness of our method is demonstrated by comparing clustering results with other methods.

DOI： 10.11234/gi1990.15.2_151

PubMed

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE COMPUTER SOC  

We propose a novel method for an exhaustive inference of gene regulatory networks from genome-wide expression data and biological knowledge. Our method performs the inferences based on module network model. In the model a module is a set of genes with similar features, and a network represents regulatory relationships among the modules. Our method makes modules using gene functional classification together with expression data. We apply our method to inferences of the networks of yeast cell-cycle. Modules inferred by our method show consistency with experimentally-determined results on yeast cell-cycle, especially on G1 phase. Robust modules built by our method permit us to infer informative regulatory relationships.

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

To elucidate fundamental constituting principle of functional modules or building blocks of metabolic networks, computational methods to analyze the network structure of metabolism are getting much attention. We propose a graph search method to extract highly conserved sub-networks of metabolic networks based on phylogenetic profile. We formulated reaction-conservation score for the measure of the phylogenetic conservation of reactions. We also formulated compound-conservation score to eliminate biologically-meaningless compounds and reduce the size of the networks. By applying our approach to the metabolic networks of 19 representative organisms selected from bacteria, archaea, and eukaryotes in the KEGG database, we detected some highly conserved sub-networks among the organisms. Comparing them to the metabolic maps in KEGG, we found they were mainly included in energy metabolism, sugar metabolism, and amino acid metabolism.

Scopus

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：COLD SPRING HARBOR LAB PRESS, PUBLICATIONS DEPT  

DOI： 10.1001/gr.1459703

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：SPRINGER-VERLAG BERLIN  

We present a method to shorten the computational time of the fluorescent DNA computing. Fluorescent DNA computing is proposed to solve intractable computation problems such as SAT problems. They use two groups of fluorescent DNA strands. One group of fluorescent DNA represents that a constraint of the given problem is satisfied, and another group represents that a constraint is unsatisfied. The calculation is executed by hybridizing them competitively to DNA beads or spots on DNA microarray. Though the biological operation used in the fluorescent DNA computing is simple, it needs the same number of beads or spots on microarray as the number of candidate solutions. In this paper, we prove that one bead or spot can represent plural candidate solutions through SAT problem, and show the algorithm and an experimental result of the fluorescent DNA computing.

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE COMPUTER SOC  

This paper presents a clustering method for the comparative analysis of genomes and metabolic pathways. Our method consists of several steps: (1) extract linear sequences of metabolic reactions (non-branching pathways) from a network of metabolic pathways, (2) perform pairwise alignments of every pair of nonbranching pathways, and (3) explore a set of genes for each aligned non-branching pathways such that those genes encode enzymes of the pathway and the genes are closely located on a genome (such as operons or gene clusters). To measure the similarity between pathways, we formalized a scoring system by using the functional hierarchy of the EC numbers of enzymes. By applying our method to the metabolic pathways in Escherichia coli, we have obtained several Pairs of pathways with similar reactions having enzymes encoded by neighbor genes on the E. coli genome. The resulting pathway pairs were mainly found in metabolisms between similar amino acids (tryptophan and histidine) and between similar sugars (fucose and rhamnose).

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：NATURE PUBLISHING GROUP  

Only a small proportion of the mouse genome is transcribed into mature messenger RNA transcripts. There is an international collaborative effort to identify all full-length mRNA transcripts from the mouse, and to ensure that each is represented in a physical collection of clones. Here we report the manual annotation of 60,770 full-length mouse complementary DNA sequences. These are clustered into 33,409 &apos;transcriptional units&apos;, contributing 90.1% of a newly established mouse transcriptome database. Of these transcriptional units, 4,258 are new protein-coding and 11,665 are new non-coding messages, indicating that non-coding RNA is a major component of the transcriptome. 41% of all transcriptional units showed evidence of alternative splicing. In protein-coding transcripts, 79% of splice variations altered the protein product. Whole-transcriptome analyses resulted in the identification of 2,431 sense-antisense pairs. The present work, completely supported by physical clones, provides the most comprehensive survey of a mammalian transcriptome so far, and is a valuable resource for functional genomics.

DOI： 10.1038/nature01266

Web of Science

researchmap

researchmap

Language：English   Publisher：IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG  

In this paper we show that the neuron filter is effective for relaxing the coefficient sensitiveness of the Hopfield neural network for combinatorial optimization problems. Since the parameters in motion equation have a significant influence on the performance of the neural network, many studies have been carried out to support determining the value of the parameters. However, not a few researchers have determined the value of the parameters experimentally yet. We show that the use of the neuron filter is effective for the parameter tuning, particularly for determining their values experimentally through simulations.

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG  

A constraint resolution scheme in the Hopfield-type neural network named "Neuron Filter" is presented for efficiently solving combinatorial optimization problems. The neuron filter produces an output that satisfies the constraints of the problem as best as possible according to both neuron inputs and outputs. This paper defines the neuron filter and shows its introduction into existing neural networks for N-queens problems and FPGA board-level routing problems. The performance is evaluated through simulations wharf the results show that our neuron filter improves the searching: capability of the neural network with the shorter computation time.

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：Scripta Technica Inc  

In this paper, we propose a neural network algorithm that uses the expanded maximum neuron model to solve the channel assignment problem of cellular radio networks, which is an NP-complete combinatorial optimization problem. The channel assignment problem demands minimizing the total interference between the assigned channels needed to satisfy all of the communication needs. The proposed expanded maximum neuron model selects multiple neurons in descending order from the neuron inputs in each neuron group. As a result, the constraints will always be satisfied for the channel assignment problem. To improve the accuracy of the solution, neuron fixing, which is a heuristic technique used in the binary neuron model, a hill-climbing term, a shaking term, and an Omega function are introduced. The effectiveness of these additions to the expanded maximum neuron model algorithm is demonstrated. Simulations of benchmark problems demonstrate the superior performance of the proposed algorithm over conventional algorithms in finding the solution.

DOI： 10.1002/(SICI)1520-6440(200011)83:11<11::AID-ECJC2>3.0.CO;2-D

Scopus

researchmap

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

This paper presents a neuron filter algorithm to satisfy two constraints of the graph-coloring problem through a separated board-level routing problem (s-BLRP) in an FPGA-based logic emulation system. For a rapid prototyping of large scale digital systems, multiple FPGA's provide an efficient logic emulation system, where signals or nets between design partitions embedded on different FPGA's are connected through crossbars. We propose a new neuron filter algorithm to satisfy the two constraints of the problem simultaneously. The simulation results in randomly generated benchmark size instances show that our neuron filter algorithm with the thinning out application provides the better routing capability with the shorter computation time.

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

In this paper, we propose an improved genetic algorithm for the traveling salesman problem (TSP) by using the `visiting order restriction theorem'. The visiting order restriction theorem gives a necessary condition for the shortest tour of TSP on the Euclidean plane by using the convex hull. The convex hull for a set of points S on a plane is defined as the smallest convex polygon that encloses S. In our method, the initial tours are produced to satisfy the necessary condition of the theorem for the shortest path without increasing the computation time. The simulation results using 10 well-known benchmark problems show that our algorithm can find better tour with shorter time than Pal's algorithm.

Scopus

researchmap

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

A constraint resolution scheme of the Hopfield neural network named «neuron mask» is presented for a class of combinatorial optimization problems. Neuron mask always satisfies constraints of selecting a solution candidate from each group so as to force the state of the neural network into a solution space. This paper presents the definition of neuron mask and the introduction into the neural network through the N-queens problem. The performance is verified by simulations on three computation modes, where neuron mask improves the performance of the neural network. © 1997 IEEE.

DOI： 10.1109/ICNN.1997.616220

Scopus

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：Springer Verlag  

A novel neural network approach using the maximum neuron model is presented for N-queens problems. The goal of the N-queens problem is to find a set of locations of N queens on an N × N chessboard such that no pair of queens commands each other. The maximum rjeuron model proposed by Takefuji et al. has been applied to two optimization problems where the optimization of objective functions is requested without constraints. This paper demonstrates the effectiveness of the maximum neuron model for constraint satisfaction problems through the N-queens problem. The performance is verified through simulations in up to 500-queens problems on the sequential mode, the N-parallel mode, and the N2-parallel mode, where our maximum neural network shows the far better performance than the existing neural networks.

DOI： 10.1007/s004220050337

Scopus

researchmap

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

複数の遺伝子間で行われている転写制御関係をグラフにより可視化したものを遺伝子制御ネットワークと呼ぶ．遺伝子制御ネットワークの全容は未だ解明されておらず，計算機によって推定することができれば医学，創薬の分野において大幅に実験コストを削減できる．その為，遺伝子制御ネットワークの推定手法の研究はバイオインフォマティクスにおいて非常に重要なテーマである．遺伝子制御ネットワークの解析手法の一つに，ベイジアンネットワークがある．このモデルは他のモデルに比べノイズに強く，データ数が少ない場合にも推定が可能である．しかしこのモデルはデータ数が増えると爆発的に探索空間が大きくなるため，探索を行うには非常に大きな計算量が必要となる．この欠点を回避するため，近似法であるグリーディ法を用いることが多い．遺伝子制御ネットワークを推定する研究では，ネットワークを推定した後に生物学実験や生物学者の議論の結果より得られた制御関係を考慮して反復的にネットワーク推定を行う．推定結果の反復構築とは，ネットワークを推定した後に新たな制御関係に応じてネットワークを修正するためにこれを考慮してネットワークを再度推定することである．しかし反復構築を行う場合，反復回数に比例して計算量が必要となる．本研究では，グリーディ法で反復的に推定結果を構築する場合における計算量の軽減のための手法を提案する．反復的な推定のために構築された結果のさらなる利用により計算量の軽減を図り，従来の反復推定の手法と比較することでその有効性を検証した．

CiNii Books

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

複数の遺伝子間で行われている転写制御関係をグラフにより可視化したものを遺伝子制御ネットワークと呼ぶ．遺伝子制御ネットワークの全容は未だ解明されておらず，計算機によって推定することができれば医学，創薬の分野において大幅に実験コストを削減できる．その為，遺伝子制御ネットワークの推定手法の研究はバイオインフォマティクスにおいて非常に重要なテーマである．遺伝子制御ネットワークの解析手法の一つに，ベイジアンネットワークがある．このモデルは他のモデルに比べノイズに強く，データ数が少ない場合にも推定が可能である．しかしこのモデルはデータ数が増えると爆発的に探索空間が大きくなるため，探索を行うには非常に大きな計算量が必要となる．この欠点を回避するため，近似法であるグリーディ法を用いることが多い．遺伝子制御ネットワークを推定する研究では，ネットワークを推定した後に生物学実験や生物学者の議論の結果より得られた制御関係を考慮して反復的にネットワーク推定を行う．推定結果の反復構築とは，ネットワークを推定した後に新たな制御関係に応じてネットワークを修正するためにこれを考慮してネットワークを再度推定することである．しかし反復構築を行う場合，反復回数に比例して計算量が必要となる．本研究では，グリーディ法で反復的に推定結果を構築する場合における計算量の軽減のための手法を提案する．反復的な推定のために構築された結果のさらなる利用により計算量の軽減を図り，従来の反復推定の手法と比較することでその有効性を検証した．

CiNii Books

researchmap

Language：English   Publisher：一般社団法人情報処理学会  

Genome analyses using short reads which are generated by the high-throughput sequencer begin by mapping reads to the target genome.Therefore, high accuracy mapping is crucial. However, conventional mapping tools cause multireads and unmapped reads, that is, the mapping accuracy is insufficient. One reason for causing those reads is that only one genome sequence is used as the reference, even if the targetorganisms are polyploidy. We propose a novel mapping method that takes alleles into account for the purpose of reducing multireads and unmapped reads, that is, improving mapping accuracy.

CiNii Books

researchmap

Language：English  

CiNii Books

researchmap

Language：Japanese  

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Gene function is researched and gene functional information which is annotated on gene is increasing continuously. Gene Set Analysis is one of a method using gene functional information, and we use it when we want to compare two groups. However, this method can not be applied to time-series gene expression profile. In this reserch, I propose a method to analyze gene function groupsand handle time-series data. The method extracts a time period in which works from a time-series gene expression profile.

CiNii Books

researchmap

Language：English   Publisher：The Institute of Electronics, Information and Communication Engineers  

Alternative splicing plays an important role in eukaryotic gene expression by producing diverse proteins from a single gene. Predicting how genes are transcribed is of great biological interest. To this end, massively parallel whole transcriptome sequencing, often referred to as RNA-Seq, is becoming widely used and is revolutionizing the cataloging isoforms using a vast number of short mRNA fragments called reads. Conventional RNA-Seq analysis methods typically align reads onto a reference genome (mapping) in order to capture the form of isoforms that each gene yields and how much of every isoform is expressed from an RNA-Seq dataset. However, a considerable number of reads cannot be mapped uniquely. Those so-called multireads that are mapped onto multiple locations due to short read length and analogous sequences inflate the uncertainty as to how genes are transcribed. This causes inaccurate gene expression estimations and leads to incorrect isoform prediction. To cope with this problem, we propose a method for isoform prediction by iterative mapping. The positions from which multireads originate can be estimated based on the information of expression levels, whereas quantification of isoform-level expression requires accurate mapping. These procedures are mutually dependent, and therefore remapping reads is essential. By iterating this cycle, our method estimates gene expression levels more precisely and hence improves predictions of alternative splicing. Our method simultaneously estimates isoform-level expressions by computing how many reads originate from each candidate isoform using an EM algorithm within a gene. To validate the effectiveness of the proposed method, we compared its performance with conventional methods using an RNA-Seq dataset derived from a human brain. The proposed method had a precision of 66.7% and outperformed conventional methods in terms of the isoform detection rate.

CiNii Books

researchmap

Language：English   Publisher：Information Processing Society of Japan (IPSJ)  

Alternative splicing plays an important role in eukaryotic gene expression by producing diverse proteins from a single gene. Predicting how genes are transcribed is of great biological interest. To this end, massively parallel whole transcriptome sequencing, often referred to as RNA-Seq, is becoming widely used and is revolutionizing the cataloging isoforms using a vast number of short mRNA fragments called reads. Conventional RNA-Seq analysis methods typically align reads onto a reference genome (mapping) in order to capture the form of isoforms that each gene yields and how much of every isoform is expressed from an RNA-Seq dataset. However, a considerable number of reads cannot be mapped uniquely. Those so-called multireads that are mapped onto multiple locations due to short read length and analogous sequences inflate the uncertainty as to how genes are transcribed. This causes inaccurate gene expression estimations and leads to incorrect isoform prediction. To cope with this problem, we propose a method for isoform prediction by iterative mapping. The positions from which multireads originate can be estimated based on the information of expression levels, whereas quantification of isoform-level expression requires accurate mapping. These procedures are mutually dependent, and therefore remapping reads is essential. By iterating this cycle, our method estimates gene expression levels more precisely and hence improves predictions of alternative splicing. Our method simultaneously estimates isoform-level expressions by computing how many reads originate from each candidate isoform using an EM algorithm within a gene. To validate the effectiveness of the proposed method, we compared its performance with conventional methods using an RNA-Seq dataset derived from a human brain. The proposed method had a precision of 66.7% and outperformed conventional methods in terms of the isoform detection rate.

CiNii Books

researchmap

Language：Japanese  

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

In bioinformatics, workflows have used frequently to combine many tools or services. But it have been difficult to use REST services by workflow tools, because there is no way to read their specification computationally. In this research, we proposed a method for using REST services inworkflows by converting REST services to SOAP services. This method classfies documents of REST services by their forms writing URL. And by using machine learning, this method extracts URLs of only REST services from classified documents. By using extracted URLs, this method generates SOAP services that access REST services. To show effectiveness of this method, we use it with example REST services registered in BioCatalogue.

CiNii Books

researchmap

researchmap

Language：Japanese  

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00068046/

Language：English  

To identify protein-protein interaction pairs with high accuracy, we propose a method for predicting these interactions based on characteristics obtained from protein-protein docking evaluations. Previous studies assumed that the required protein affinity strength for an interaction was not dependent on protein functions. However, the protein affinity strength appears to differ with different docking schemes, such as rigid-body or flexible docking, and these schemes may be related to protein functions. Thus, we propose a new scoring system that is based on statistical analysis of affinity score distributions sampled by their protein functions. As a result, of all possible protein pair combinations, a newly developed method improved prediction accuracy of F-measures. In particular, for bound antibody-antigen pairs, we obtained 50.0% recall (= sensitivity) with higher F-measures compared with previous studies. In addition, by combining two proposed scoring systems, Receptor-Focused Z-scoring and Ligand-Focused Z-scoring, further improvement was achieved. This result suggested that the proposed prediction method improved the prediction accuracy (i.e., F-measure), with few false positives, by taking biological functions of protein pairs into consideration. © 2010 Information Processing Society of Japan.

DOI： 10.2197/ipsjtbio.3.10

Scopus

researchmap

researchmap

Language：English  

To identify protein-protein interaction pairs with high accuracy, we propose a method for predicting these interactions based on characteristics obtained from protein-protein docking evaluations. Previous studies assumed that the required protein affinity strength for an interaction was not dependent on protein functions. However, the protein affinity strength appears to differ with different docking schemes, such as rigid-body or flexible docking, and these schemes may be related to protein functions. Thus, we propose a new scoring system that is based on statistical analysis of affinity score distributions sampled by their protein functions. As a result, of all possible protein pair combinations, a newly developed method improved prediction accuracy of F-measures. In particular, for bound antibody-antigen pairs, we obtained 50.0% recall (= sensitivity) with higher F-measures compared with previous studies. In addition, by combining two proposed scoring systems, Receptor-Focused Z-scoring and Ligand-Focused Z-scoring, further improvement was achieved. This result suggested that the proposed prediction method improved the prediction accuracy (i.e., F-measure), with few false positives, by taking biological functions of protein pairs into consideration. © 2010 Information Processing Society of Japan.

DOI： 10.2197/ipsjtbio.3.10

Scopus

researchmap

Language：Japanese  

CiNii Books

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

The amount of compounds in public databases goes on increasing. It is necessary not only to accumulate compound data but also to classify them such as clustering based on their structures, because their activities on proteins depend on the structures. It is necessary to calculate the similarity between compounds, but as the number of compounds increases, the calculation time increases. In this research, we propose a method for reducing the number of calculating similarity based on the property of the Tanimoto coefficient, and show the effectiveness by the experiment with compound data.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00058900/

researchmap

researchmap

Language：English   Publisher：SPRINGER HEIDELBERG  

Gene expressions differ depending on tissue types and developmental stages. Analyzing how each gene is expressed is thus important. One way of analyzing gene expression patterns is to identify tissue-specific functions. This is useful for understanding how vital activities are performed. DNA microarray has been widely used to observe gene expressions exhaustively. However, comparing the expression value of a gene to that of other genes is impossible, as the gene expression value of a condition is measured as a proportion of that for the same gene under a control condition. We therefore could not determine whether one gene is more expressed than other genes. Cap analysis gene expression (CAGE) allows high-throughput analysis of gene expressions by counting the number of cDNAs of expressed genes. CAGE enables comparison of the expression value of the gene to that of other genes in the same tissue. In this study, we propose a method for exploring tissue-specific functions using data from CAGE. To identify tissue-specificity, one of the simplest ways is to assume that the function of the most expressed gene is regarded as the most tissue-specific. However, the most expressed gene in a tissue might highly express in all tissues, as seen with housekeeping genes. Functions of such genes cannot be tissue-specific. To remove these from consideration, we propose measuring tissue specificity of functions based on information content of gene ontology terms. We applied our method to data from 16 human tissues and 22 mouse tissues. The results from liver and prostate gland indicated that well-known functions of these tissues, such as functions related to signaling and muscle in prostate gland and immune function in liver, displayed high rank.

DOI： 10.1007/s11517-007-0274-y

Web of Science

researchmap

Language：English   Publisher：SPRINGER HEIDELBERG  

Gene expressions differ depending on tissue types and developmental stages. Analyzing how each gene is expressed is thus important. One way of analyzing gene expression patterns is to identify tissue-specific functions. This is useful for understanding how vital activities are performed. DNA microarray has been widely used to observe gene expressions exhaustively. However, comparing the expression value of a gene to that of other genes is impossible, as the gene expression value of a condition is measured as a proportion of that for the same gene under a control condition. We therefore could not determine whether one gene is more expressed than other genes. Cap analysis gene expression (CAGE) allows high-throughput analysis of gene expressions by counting the number of cDNAs of expressed genes. CAGE enables comparison of the expression value of the gene to that of other genes in the same tissue. In this study, we propose a method for exploring tissue-specific functions using data from CAGE. To identify tissue-specificity, one of the simplest ways is to assume that the function of the most expressed gene is regarded as the most tissue-specific. However, the most expressed gene in a tissue might highly express in all tissues, as seen with housekeeping genes. Functions of such genes cannot be tissue-specific. To remove these from consideration, we propose measuring tissue specificity of functions based on information content of gene ontology terms. We applied our method to data from 16 human tissues and 22 mouse tissues. The results from liver and prostate gland indicated that well-known functions of these tissues, such as functions related to signaling and muscle in prostate gland and immune function in liver, displayed high rank.

DOI： 10.1007/s11517-007-0274-y

Web of Science

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Finding a common pattern among metabolic network gives important information on the evolution of the pathway. In previous works, it has been done to define the similarity of the reactions in the metabolic network by using the EC number. There are many reactions whose conformational changes of the compounds are similar, though their EC numbers are quite different. In this research, we propose a method for extracting the similarity in the network by using Reaction Classification Number.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00058990/

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

This article presents a method to extract tissue-specific metabolic-pathways from absolute expression data based on outlier detection. By mapping tissue-specific genes to metabolic pathways, majority of acquired pathways are in flagments. It is caused by existance of nontissue-specific genes between small fragments, which are considered to work with a unit. We defined a score from expression profile of a pathway and gene which are included in the pathway?

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00058988/

Language：English   Publisher：SPRINGER-VERLAG BERLIN  

We propose a computational inference of transcriptional regulatory networks from transcription start sites identified by Cap Analysis of Gene Expression (CAGE). Inference of the regulatory networks is a challenging task, and is performed by analyzing observed dependencies of transcription levels. Binding of transcription factors to an upstream sequence of a gene enables transcription initiation from a specific genomic position. The position is called Transcription Start Site (TSS). Eukaryotes have multiple TSSs for a single gene, which reflect diversity in combinatorial regulation by transcription factors. By the CAGE high-throughput profiling genome-wide identification of activated TSSs under various experimental conditions yields CAGE transcriptome datasets. To distinguish transcription responses caused by the diversity in combinatorial regulations, inference that takes multiple TSSs into account will be more effective than the conventional ones using only transcription levels of microarray dataset.
In this paper we report a feasibility study of inference of transcriptional regulatory network by using CAGE dataset. This is a preliminary study for inference that takes multiple TSSs into account. To perform inference of the regulatory networks, we used Bayesian network model which is appropriate to represent distribution of TSSs observed in CAGE dataset. Using the CAGE dataset published from the FANTOM3 project in RIKEN, we applied the model to inference of the regulatory networks of Mus musculus. For the purpose of the feasibility study, we compared inferred networks from the CAGE dataset with those from microarray dataset. Based on the comparative analysis, we confirmed that network inference by using the CAGE dataset is feasible like the case of conventional inferences by using microarray dataset. Based on the confirmation we discussed what we can reveal through the inference that takes into account multiple TSSs identified by a CAGE dataset.

Web of Science

researchmap

Language：English   Publisher：SPRINGER-VERLAG BERLIN  

We propose a computational inference of transcriptional regulatory networks from transcription start sites identified by Cap Analysis of Gene Expression (CAGE). Inference of the regulatory networks is a challenging task, and is performed by analyzing observed dependencies of transcription levels. Binding of transcription factors to an upstream sequence of a gene enables transcription initiation from a specific genomic position. The position is called Transcription Start Site (TSS). Eukaryotes have multiple TSSs for a single gene, which reflect diversity in combinatorial regulation by transcription factors. By the CAGE high-throughput profiling genome-wide identification of activated TSSs under various experimental conditions yields CAGE transcriptome datasets. To distinguish transcription responses caused by the diversity in combinatorial regulations, inference that takes multiple TSSs into account will be more effective than the conventional ones using only transcription levels of microarray dataset.
In this paper we report a feasibility study of inference of transcriptional regulatory network by using CAGE dataset. This is a preliminary study for inference that takes multiple TSSs into account. To perform inference of the regulatory networks, we used Bayesian network model which is appropriate to represent distribution of TSSs observed in CAGE dataset. Using the CAGE dataset published from the FANTOM3 project in RIKEN, we applied the model to inference of the regulatory networks of Mus musculus. For the purpose of the feasibility study, we compared inferred networks from the CAGE dataset with those from microarray dataset. Based on the comparative analysis, we confirmed that network inference by using the CAGE dataset is feasible like the case of conventional inferences by using microarray dataset. Based on the confirmation we discussed what we can reveal through the inference that takes into account multiple TSSs identified by a CAGE dataset.

Web of Science

researchmap

Language：English   Publisher：PUBLIC LIBRARY SCIENCE  

T he international FANTOM consortium aims to produce a comprehensive picture of the mammalian transcriptome, based upon an extensive cDNA collection and functional annotation of full-length enriched cDNAs. The previous dataset, FANTOM(2), comprised 60,770 full- length enriched cDNAs. Functional annotation revealed that this cDNA dataset contained only about half of the estimated number of mouse protein- coding genes, indicating that a number of cDNAs still remained to be collected and identified. To pursue the complete gene catalog that covers all predicted mouse genes, cloning and sequencing of full- length enriched cDNAs has been continued since FANTOM2. In FANTOM3, 42,031 newly isolated cDNAs were subjected to functional annotation, and the annotation of 4,347 FANTOM2 cDNAs was updated. To accomplish accurate functional annotation, we improved our automated annotation pipeline by introducing new coding sequence prediction programs and developed a Web- based annotation interface for simplifying the annotation procedures to reduce manual annotation errors. Automated coding sequence and function prediction was followed with manual curation and review by expert curators. A total of 102,801 full- length enriched mouse cDNAs were annotated. Out of 102,801 transcripts, 56,722 were functionally annotated as protein coding ( including partial or truncated transcripts), providing to our knowledge the greatest current coverage of the mouse proteome by full- length cDNAs. The total number of distinct non- protein- coding transcripts increased to 34,030. The FANTOM3 annotation system, consisting of automated computational prediction, manual curation, and. nal expert curation, facilitated the comprehensive characterization of the mouse transcriptome, and could be applied to the transcriptomes of other species.

DOI： 10.1371/journal.pgen.0020062

Web of Science

researchmap

Language：English   Publisher：PUBLIC LIBRARY SCIENCE  

T he international FANTOM consortium aims to produce a comprehensive picture of the mammalian transcriptome, based upon an extensive cDNA collection and functional annotation of full-length enriched cDNAs. The previous dataset, FANTOM(2), comprised 60,770 full- length enriched cDNAs. Functional annotation revealed that this cDNA dataset contained only about half of the estimated number of mouse protein- coding genes, indicating that a number of cDNAs still remained to be collected and identified. To pursue the complete gene catalog that covers all predicted mouse genes, cloning and sequencing of full- length enriched cDNAs has been continued since FANTOM2. In FANTOM3, 42,031 newly isolated cDNAs were subjected to functional annotation, and the annotation of 4,347 FANTOM2 cDNAs was updated. To accomplish accurate functional annotation, we improved our automated annotation pipeline by introducing new coding sequence prediction programs and developed a Web- based annotation interface for simplifying the annotation procedures to reduce manual annotation errors. Automated coding sequence and function prediction was followed with manual curation and review by expert curators. A total of 102,801 full- length enriched mouse cDNAs were annotated. Out of 102,801 transcripts, 56,722 were functionally annotated as protein coding ( including partial or truncated transcripts), providing to our knowledge the greatest current coverage of the mouse proteome by full- length cDNAs. The total number of distinct non- protein- coding transcripts increased to 34,030. The FANTOM3 annotation system, consisting of automated computational prediction, manual curation, and. nal expert curation, facilitated the comprehensive characterization of the mouse transcriptome, and could be applied to the transcriptomes of other species.

DOI： 10.1371/journal.pgen.0020062

Web of Science

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Books

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

The accumulation of gene-expression profiles can allow an inference of a gene regulatory network by using a profile measured under a number of experimental conditions. However, in case of applying to such profile, the conventional methods such as module network model may not perform an inference accurately enough, because of following two facts. 1) Module network can accurately perform an inference only for regulated genes that show similar gene-expression patterns under almost all experimental conditions. 2) In a gene-expression profile that includes more conditions, fewer genes show similar gene-expression patterns. To alleviate the accuracy loss, we utilized a biclustering result for an inference. We performed an inference for regulated genes that were included in a detected bicluster by using gene-expression patterns only under experimental conditions included in the bicluster. We demonstrate the effectiveness of our method by applying to inferences of gene regulatory networks by using various gene-expression profiles of budding yeast.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00017157/

researchmap

researchmap

researchmap

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

This article presents a method to extract metabolic sub-networks conserved among organisms based on similarity between phylogenetic profiles of enzymatic reactions. We formulated a profile similarity score based on the phylogenetic profile. By using the score, we perform a connected component search algorithm. By applying our approach to the metabolic networks of 19 representative organisms selected from bacteria, archaea, and eukaryotes in the KEGG database, we detected some highly conserved sub-networks among the organisms. We obtained one of hte sub-network conserved among most of the organisms and the tryptophan biosynthesis pathway conserved between bacteria and archaea.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00059080/

researchmap

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

The gene regulation networks are directed graph expression of gene regulations. Because the amount of expression profile's samples is rarely enough to robustly learn dependencies of a large number of genes, any methods have ot succeeded in a precise inference of gene regulation network. In this research a gene regulation network is infered by modeling to Module Bayesian Networks, which are introduced the notion of module in which variables set has the same dependency, and a module scoring method is proposed so that infered module's genes have the similar annotation of prior knowledge. A performance of proposed method is evaluated on the actual gene expression profile.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00033354/

Language：English  

researchmap

Language：English  

researchmap

Language：English   Publisher：SPRINGER-VERLAG BERLIN  

We present a method to shorten the computational time of the fluorescent DNA computing. Fluorescent DNA computing is proposed to solve intractable computation problems such as SAT problems. They use two groups of fluorescent DNA strands. One group of fluorescent DNA represents that a constraint of the given problem is satisfied, and another group represents that a constraint is unsatisfied. The calculation is executed by hybridizing them competitively to DNA beads or spots on DNA microarray. Though the biological operation used in the fluorescent DNA computing is simple, it needs the same number of beads or spots on microarray as the number of candidate solutions. In this paper, we prove that one bead or spot can represent plural candidate solutions through SAT problem, and show the algorithm and an experimental result of the fluorescent DNA computing.

Web of Science

researchmap

Language：English   Publisher：NATURE PUBLISHING GROUP  

Only a small proportion of the mouse genome is transcribed into mature messenger RNA transcripts. There is an international collaborative effort to identify all full-length mRNA transcripts from the mouse, and to ensure that each is represented in a physical collection of clones. Here we report the manual annotation of 60,770 full-length mouse complementary DNA sequences. These are clustered into 33,409 &apos;transcriptional units&apos;, contributing 90.1% of a newly established mouse transcriptome database. Of these transcriptional units, 4,258 are new protein-coding and 11,665 are new non-coding messages, indicating that non-coding RNA is a major component of the transcriptome. 41% of all transcriptional units showed evidence of alternative splicing. In protein-coding transcripts, 79% of splice variations altered the protein product. Whole-transcriptome analyses resulted in the identification of 2,431 sense-antisense pairs. The present work, completely supported by physical clones, provides the most comprehensive survey of a mammalian transcriptome so far, and is a valuable resource for functional genomics.

DOI： 10.1038/nature01266

Web of Science

researchmap

researchmap

Language：English   Publishing type：Rapid communication, short report, research note, etc. (scientific journal)   Publisher：IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG  

In this paper we show that the neuron filter is effective for relaxing the coefficient sensitiveness of the Hopfield neural network for combinatorial optimization problems. Since the parameters in motion equation have a significant influence on the performance of the neural network, many studies have been carried out to support determining the value of the parameters. However, not a few researchers have determined the value of the parameters experimentally yet. We show that the use of the neuron filter is effective for the parameter tuning, particularly for determining their values experimentally through simulations.

Web of Science

researchmap

Language：English   Publisher：IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG  

A constraint resolution scheme in the Hopfield-type neural network named "Neuron Filter" is presented for efficiently solving combinatorial optimization problems. The neuron filter produces an output that satisfies the constraints of the problem as best as possible according to both neuron inputs and outputs. This paper defines the neuron filter and shows its introduction into existing neural networks for N-queens problems and FPGA board-level routing problems. The performance is evaluated through simulations wharf the results show that our neuron filter improves the searching: capability of the neural network with the shorter computation time.

Web of Science

researchmap

researchmap

researchmap

Language：English   Publisher：The Institute of Electronics, Information and Communication Engineers  

In this paper, we propose a non-feedback neuron filter algorithm in the neural network. for the board-level routing problem (BLRP) in an FPGA-based logic emulation system. For a rapid prototyping of large scale digital systems, multiple FPGA's provide an efficient logic emulation system, where signals or nets between design partitions embedded on different FPGA's are connected through crossbars. We propose a new neuron filter algorithm to satisfy the two constraints of the problem simultaneously. The simulation results in randomly generated benchmark size instances show that our neuron filter algorithm with the thinning out application provides the better routing capability with the shorter computation time.

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

In this paper, we propose an application of the "visiting order restriction theorem" to Pal's algorithm, a genetic algorithm for the traveling salesman problem (TSP). The visiting order restriction theorem gives a necessary condition for the shortest tour of TSP on the Euclidean plane by using the convex hull. The convex hull of a set of points S on a plane is defined as the smallest convex polygon that encloses S. The O (|S|log|S|)-time algorithm has been known to find the convex hull of a set of points S. The genetic algorithm, which simulates the process of the natural selection, is known as one of best approximate algorithms for TSP. In our method, the initial tours are produced to satisfy the necessary condeition for the shortest path in order to improve the solution quality. The simulation results using benchmark problems show our genetic algorithm finds shorter tours than P'al's algorithm at the large size benchmark problems.

CiNii Books

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

The maximum neuron model provides efficient neural network solutions for combinatorial optimization problems. In this "winner-take-all" model, one and only one neuron with the maximum input value is always fired in each group of neurons to satisfy the selection constraint. The maximum neuron model can not only limit the searching space, but also reduce the computation load. In this paper, we propose two methods for selecting one neuron among two or more neurons which have the same maximum input value, named "least index method" and "previous selection method". The simulation results in N-queens problems show that the latter method performs better than the former method generally, and that the previous selection method on the N-parallel mode provides the suitable algorithm for hardware implementation of large-scale neural networks.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00013273/

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

The maximum neuron model provides efficient neural network solutions for combinatorial optimization problems. In this "winner-take-all" model, one and only one neuron with the maximum input value is always fired in each group of neurons to satisfy the selection constraint. The maximum neuron model can not only limit the searching space, but also reduce the computation load. In this paper, we propose two methods for selecting one neuron among two or more neurons which have the same maximum input value, named "least index method" and "previous selection method". The simulation results in N-queens problems show that the latter method performs better than the former method generally, and that the previous selection method on the N-parallel mode provides the suitable algorithm for hardware implementation of large-scale neural networks.

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00013273/

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

The maximum neuron model provides efficient neural network solutions for combinatorial optimization problems. In this "winner-take-all" model, one and only one neuron with the maximum input value is always fired in each group of neurons to satisfy the selection constrain. The maximum neuron model can not only reduce the searching space, but also save the computation load. In this paper, we propose two methods of selecting one neuron from two or more neurons which have same maximum input value, named "Least index method" and "Previous selection method". The simulation results show that the performance of the previous selection method performs better than the least index method generally.

CiNii Books

researchmap

researchmap

researchmap

Event date： 2018.11

Venue：軽井沢  

researchmap

Event date： 2018.11

Venue：Phoenix, AZ, USA  

In high-dimensional regression, where the number of explanatory variables is much larger than the sample size, a statistical problem known as the ‘curse of dimensionality’ arises. The strategy of performing all possible regressions is computationally-impractical. We model using non-convex discrete optimization to minimize MSE. We describe an application involving a large number of SNPs in genomic studies for cancer detection.

researchmap

Event date： 2018.11

researchmap

Event date： 2018.7

researchmap

Event date： 2018.5

Venue：札幌  

researchmap

Event date： 2018.1

Venue：大阪大学  

researchmap

Event date： 2018.1

researchmap

Event date： 2017.10

Venue：札幌  

researchmap

Event date： 2017.8

researchmap

Event date： 2017.5

Venue：仙台  

researchmap

Event date： 2017.5

Venue：仙台  

researchmap

Event date： 2017.2

researchmap

Event date： 2017.2

researchmap

Event date： 2016.11

Venue：Splendor Hotel, Taichung, Taiwan  

researchmap

Event date： 2016.7

Venue：Swan and Dolphin Hotel, Orlando, Florida, USA  

researchmap

Event date： 2016.7

Venue：Swan and Dolphin Hotel, Orlando, Florida, USA  

researchmap

Event date： 2016.6

Venue：大阪大学銀杏会館, 大阪  

researchmap

Event date： 2016.6

Venue：大阪大学銀杏会館, 大阪  

researchmap

Event date： 2016.4

researchmap

Event date： 2016.4

researchmap

Event date： 2016.3

researchmap

Event date： 2016.3

researchmap

Event date： 2016.3

researchmap

Event date： 2016.3

researchmap

Event date： 2016.3

researchmap

Event date： 2015.11

researchmap

Event date： 2015.10

researchmap

Event date： 2015.9

researchmap

Event date： 2015.7

researchmap

Event date： 2015.7

researchmap

Event date： 2015.6

researchmap

Event date： 2015.6

researchmap

Event date： 2015.6

researchmap

Event date： 2015.3

researchmap

Event date： 2015.3

researchmap

Event date： 2014.11

researchmap

Event date： 2014.10

researchmap

Event date： 2014.10

researchmap

Event date： 2014.9

researchmap

Event date： 2014.9

researchmap

Event date： 2014.9

researchmap

Event date： 2014.7

researchmap

Event date： 2014.6

researchmap

Event date： 2014.6

researchmap

Event date： 2014.3

researchmap

Event date： 2014.3

researchmap

Event date： 2013.12

researchmap

Event date： 2013.11

researchmap

Event date： 2013.10

researchmap

Event date： 2013.9

researchmap

Event date： 2013.9

researchmap

Event date： 2013.9

researchmap

Event date： 2013.8

researchmap

Event date： 2013.7

researchmap

Event date： 2013.7

Short Papers, 3225

researchmap

Event date： 2013.7

researchmap

Event date： 2013.7

researchmap

Event date： 2013.7

researchmap

Event date： 2013.7

Short Papers, 3055

researchmap

Event date： 2013.3

researchmap

Event date： 2013.2

researchmap

Event date： 2013.2

researchmap

Event date： 2012.12

researchmap

Event date： 2012.12

researchmap

Event date： 2012.12

researchmap

Event date： 2012.12

researchmap

Event date： 2012.12

researchmap

Event date： 2012.11

researchmap

Event date： 2012.10

researchmap

Event date： 2012.10

researchmap

Event date： 2012.10

researchmap

Event date： 2012.10

researchmap

Event date： 2012.10

researchmap

Event date： 2012.10

researchmap

Event date： 2012.7

researchmap

Event date： 2012.6

researchmap

Event date： 2012.6

researchmap

Event date： 2012.5

researchmap

Event date： 2012.3

researchmap

Event date： 2012.1

researchmap

Event date： 2011.12

researchmap

Event date： 2011.12

researchmap

Event date： 2011.11

researchmap

Event date： 2011.11

researchmap

Event date： 2011.11

researchmap

Event date： 2011.11

researchmap

Event date： 2011.11

researchmap

Event date： 2011.11

researchmap

Event date： 2011.7

researchmap

Event date： 2011.7

researchmap

Event date： 2011.7

researchmap

Event date： 2011.7

researchmap

Event date： 2011.7

researchmap

Event date： 2011.6

researchmap

Event date： 2011.3

researchmap

Event date： 2011.3

researchmap

Event date： 2010.12

researchmap

Event date： 2010.12

researchmap

Event date： 2010.12

researchmap

Event date： 2010.9

researchmap

Event date： 2010.7

researchmap

Event date： 2010.7

researchmap

Event date： 2010.7

researchmap

Event date： 2010.7

researchmap

Event date： 2010.3

researchmap

Event date： 2010.3

researchmap

Event date： 2009.12

researchmap

Event date： 2009.12

researchmap

Event date： 2009.12

researchmap

Event date： 2009.6

researchmap

Event date： 2009.6

researchmap

Event date： 2009.3

researchmap

Event date： 2008.12

researchmap

Event date： 2008.12

researchmap

Event date： 2008.12

researchmap

Event date： 2008.12

researchmap

Event date： 2008.12

researchmap

Event date： 2008.12

researchmap

Event date： 2008.12

researchmap

Event date： 2008.9

researchmap

Event date： 2008.7

researchmap

Event date： 2008.7

researchmap

Event date： 2008.7

researchmap

Event date： 2008.7

researchmap

Event date： 2008.5

researchmap

Event date： 2008.5

researchmap

Event date： 2008.3

researchmap

Event date： 2008.3

researchmap

Event date： 2007.12

researchmap

Event date： 2007.12

researchmap

Event date： 2007.12

researchmap

Event date： 2007.12

researchmap

Event date： 2007.9

researchmap

Event date： 2007.3

researchmap

Event date： 2007.3

researchmap

Event date： 2007.3

researchmap

Event date： 2006.12

researchmap

Event date： 2006.11

researchmap

Event date： 2006.8

researchmap

Event date： 2006.8

researchmap

Event date： 2006.7

researchmap

Event date： 2006.6

researchmap

Event date： 2006.3

researchmap

Event date： 2006.3

researchmap

Event date： 2005.12

researchmap

Event date： 2005.12

researchmap

Event date： 2005.12

researchmap

Event date： 2005.12

researchmap

Event date： 2005.12

researchmap

Event date： 2005.12

researchmap

Event date： 2005.5

researchmap

Event date： 2005.4

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.12

researchmap

Event date： 2004.9

researchmap

Event date： 2004.8

researchmap

Event date： 2004.8

researchmap

Event date： 2003.12

researchmap

Event date： 2003.12

researchmap

Event date： 2003.12

researchmap

Event date： 2003.12

researchmap

Event date： 2003.12

researchmap

Event date： 2003.12

researchmap

Event date： 2003.12

researchmap

Event date： 2003.12

researchmap

Event date： 2003.10

researchmap

Event date： 2003.9

researchmap

Event date： 2003.9

researchmap

Event date： 2003.7

researchmap

Event date： 2003.7

researchmap

Event date： 2003.3

researchmap

Event date： 2002.12

researchmap

Event date： 2002.12

researchmap

Event date： 2002.12

researchmap

Event date： 2002.12

researchmap

Event date： 2002.12

researchmap

Event date： 2002.12

researchmap

Event date： 2002.12

researchmap

Event date： 2002.6

researchmap

Event date： 2002.5

researchmap

Event date： 2002.5

researchmap

Event date： 2002.4

researchmap

Event date： 2001.12

researchmap

Event date： 2001.12

researchmap

Event date： 2001.6

researchmap

Event date： 2001.3

researchmap

Event date： 2001.3

researchmap

Event date： 2000.12

researchmap

Event date： 2000.12

researchmap

Event date： 2000.7

researchmap

Event date： 1999.7

researchmap

Event date： 1999.7

researchmap

Event date： 1998.12

researchmap

Event date： 1998.10

researchmap

Event date： 1998.3

researchmap

Event date： 1998.1

researchmap

Event date： 1997.6

researchmap

Event date： 1997.3

researchmap

Event date： 1996.11

researchmap

Event date： 1996.1

researchmap

researchmap