Researchers need a rich trove of genomic datasets that they can leverage to gain a better understanding of the genetic basis of the human genome and identify associations between phenol-types and specific parts of DNA. However, sharing genomic datasets that include sensitive genetic or medical information of individuals can lead to serious privacy-related consequences if data lands in the wrong hands. Restricting access to genomic datasets is one solution, but this greatly reduces their usefulness for research purposes. To allow sharing of genomic datasets while addressing these privacy concerns, several studies propose privacy-preserving mechanisms for data sharing. Differential privacy is one of such mechanisms that formalize rigorous mathematical foundations to provide privacy guarantees while sharing aggregated statistical information about a dataset. Nevertheless, it has been shown that the original privacy guarantees of DP-based solutions degrade when there are dependent tuples in the dataset, which is a common scenario for genomic datasets (due to the existence of family members). In this work, we introduce a new mechanism to mitigate the vulnerabilities of the inference attacks on differentially private query results from genomic datasets including dependent tuples. We propose a utility-maximizing and privacy-preserving approach for sharing statistics by hiding selective SNPs of the family members as they participate in a genomic dataset. By evaluating our mechanism on a real-world genomic dataset, we empirically demonstrate that our proposed mechanism can achieve up to 40% better privacy than state-of-the-art DP-based solutions, while near-optimally minimizing utility loss.
{"title":"Tuning Privacy-Utility Tradeoff in Genomic Studies Using Selective SNP Hiding.","authors":"Nour Almadhoun Alserr, Gulce Kale, Onur Mutlu, Oznur Tastan, Erman Ayday","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>Researchers need a rich trove of genomic datasets that they can leverage to gain a better understanding of the genetic basis of the human genome and identify associations between phenol-types and specific parts of DNA. However, sharing genomic datasets that include sensitive genetic or medical information of individuals can lead to serious privacy-related consequences if data lands in the wrong hands. Restricting access to genomic datasets is one solution, but this greatly reduces their usefulness for research purposes. To allow sharing of genomic datasets while addressing these privacy concerns, several studies propose privacy-preserving mechanisms for data sharing. Differential privacy is one of such mechanisms that formalize rigorous mathematical foundations to provide privacy guarantees while sharing aggregated statistical information about a dataset. Nevertheless, it has been shown that the original privacy guarantees of DP-based solutions degrade when there are dependent tuples in the dataset, which is a common scenario for genomic datasets (due to the existence of family members). In this work, we introduce a new mechanism to mitigate the vulnerabilities of the inference attacks on differentially private query results from genomic datasets including dependent tuples. We propose a utility-maximizing and privacy-preserving approach for sharing statistics by hiding selective SNPs of the family members as they participate in a genomic dataset. By evaluating our mechanism on a real-world genomic dataset, we empirically demonstrate that our proposed mechanism can achieve up to 40% better privacy than state-of-the-art DP-based solutions, while near-optimally minimizing utility loss.</p>","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"2023 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10306260/pdf/nihms-1902817.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"9742812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Is bioinformatics a discipline? Perhaps a more fundamental question is what is a discipline? According to the Webster's New World Dictionary it is "a system of rules, as for a monastic order." A little harsh perhaps. A more apropos definition is a field of study that stands alone, yet at the same time can be integrated into a large picture of human understanding. In this sense is bioinformatics a discipline? Certainly a question worthy of debate. Most of this audience would say yes, others would say it is a part of computer or information science, others a methodology within the biological sciences and others something destined to merge with medical informatics. If we are struggling for a unique definition perhaps this indicates we want it to be something that stands alone and distinct? I will address this question.Regardless of how we classify bioinformatics, we are dealing with a fledgling enterprise that arose out of the human genome project and has become an interpreter of the genomic language of DNA and is attempting to decipher the more complex languages where proteins are the nouns and interactions the syntax and pathways the sentences and living systems the complete volume. All fledglings learn how to discover, how to grow and adapt and how to live in a complex world. Bioinformatics is no exception. We will explore where we are on this learning curve with examples from our work and others. An exploration covering, methods development, putting the "bio" back in bioinformatics, quality control, and making the leap to systems biology.
{"title":"The Future of Bioinformatics","authors":"P. Bourne","doi":"10.1126/article.63850","DOIUrl":"https://doi.org/10.1126/article.63850","url":null,"abstract":"Is bioinformatics a discipline? Perhaps a more fundamental question is what is a discipline? According to the Webster's New World Dictionary it is \"a system of rules, as for a monastic order.\" A little harsh perhaps. A more apropos definition is a field of study that stands alone, yet at the same time can be integrated into a large picture of human understanding. In this sense is bioinformatics a discipline? Certainly a question worthy of debate. Most of this audience would say yes, others would say it is a part of computer or information science, others a methodology within the biological sciences and others something destined to merge with medical informatics. If we are struggling for a unique definition perhaps this indicates we want it to be something that stands alone and distinct? I will address this question.Regardless of how we classify bioinformatics, we are dealing with a fledgling enterprise that arose out of the human genome project and has become an interpreter of the genomic language of DNA and is attempting to decipher the more complex languages where proteins are the nouns and interactions the syntax and pathways the sentences and living systems the complete volume. All fledglings learn how to discover, how to grow and adapt and how to live in a complex world. Bioinformatics is no exception. We will explore where we are on this learning curve with examples from our work and others. An exploration covering, methods development, putting the \"bio\" back in bioinformatics, quality control, and making the leap to systems biology.","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"24 1","pages":"3"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82758888","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-01-01DOI: 10.1142/9781848161092_0005
M. Bodén
The intra-nuclear organisation of proteins is based on possibly transient interactions with morphologically defined compartments like the nucleolus. The fluidity of trafficking challenges the development of models that accurately identify compartment membership for novel proteins. A growing inventory of nucleolar proteins is here used to train a support-vector machine to recognise sequence features that allow the automatic assignment of compartment membership. We explore a range of sequence-kernels and find that while some success is achieved with a profile-based local alignment kernel, the problem is ill-suited to a standard compartment-classification approach.
{"title":"Predicting Nucleolar Proteins Using Support-Vector Machines","authors":"M. Bodén","doi":"10.1142/9781848161092_0005","DOIUrl":"https://doi.org/10.1142/9781848161092_0005","url":null,"abstract":"The intra-nuclear organisation of proteins is based on possibly transient interactions with morphologically defined compartments like the nucleolus. The fluidity of trafficking challenges the development of models that accurately identify compartment membership for novel proteins. A growing inventory of nucleolar proteins is here used to train a support-vector machine to recognise sequence features that allow the automatic assignment of compartment membership. We explore a range of sequence-kernels and find that while some success is achieved with a profile-based local alignment kernel, the problem is ill-suited to a standard compartment-classification approach.","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"248 1","pages":"19-28"},"PeriodicalIF":0.0,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80655398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-01-01DOI: 10.1142/9781848161092_0017
Melissa J. Davis, A. Newman, I. Khan, J. Hunter, M. Ragan
Many efforts at standardising terminologies within the biological domain have resulted in the construction of hierarchical controlled vocabularies that capture domain knowledge. Vocabularies, such as the PSI-MI vocabulary, capture both deep and extensive domain knowledge, in the OBO (Open Biomedical Ontologies) format. However hierarchical vocabularies, such as PSI-MI which are represented in OBO, only represent simple parent-child relationships between terms. By contrast, ontologies constructed using the Web Ontology Language (OWL), such as BioPax, define many richer types of relationships between terms. OWL provides a semantically rich structured language for expressing classes and sub-classes of entities and properties, relationships between them and domain-specific rules or axioms that can be applied to extract new information through semantic inferencing. In order to fully exploit the domain knowledge inherent in domain-specific controlled vocabularies, they need to be represented as OWL-DL ontologies, rather than in formats such as OBO. In this paper, we describe a method for converting OBO vocabularies into OWL and class instances represented as OWL-RDF triples. This approach preserves the hierarchical arrangement of the domain knowledge whilst also making the underlying parent-child relationships available to inferencing engines. This approach also has clear advantages over existing methods which incorporate terms from external controlled vocabularies as literals stripped of the context associated with their place in the hierarchy. By preserving this context, we enable machine inferencing over the ordered domain knowledge captured in OBO controlled vocabularies
{"title":"Integrating Hierarchical Controlled Vocabularies With OWL Ontology: A Case Study from the Domain of Molecular Interactions","authors":"Melissa J. Davis, A. Newman, I. Khan, J. Hunter, M. Ragan","doi":"10.1142/9781848161092_0017","DOIUrl":"https://doi.org/10.1142/9781848161092_0017","url":null,"abstract":"Many efforts at standardising terminologies within the biological domain have resulted in the construction of hierarchical controlled vocabularies that capture domain knowledge. Vocabularies, such as the PSI-MI vocabulary, capture both deep and extensive domain knowledge, in the OBO (Open Biomedical Ontologies) format. However hierarchical vocabularies, such as PSI-MI which are represented in OBO, only represent simple parent-child relationships between terms. By contrast, ontologies constructed using the Web Ontology Language (OWL), such as BioPax, define many richer types of relationships between terms. OWL provides a semantically rich structured language for expressing classes and sub-classes of entities and properties, relationships between them and domain-specific rules or axioms that can be applied to extract new information through semantic inferencing. In order to fully exploit the domain knowledge inherent in domain-specific controlled vocabularies, they need to be represented as OWL-DL ontologies, rather than in formats such as OBO. In this paper, we describe a method for converting OBO vocabularies into OWL and class instances represented as OWL-RDF triples. This approach preserves the hierarchical arrangement of the domain knowledge whilst also making the underlying parent-child relationships available to inferencing engines. This approach also has clear advantages over existing methods which incorporate terms from external controlled vocabularies as literals stripped of the context associated with their place in the hierarchy. By preserving this context, we enable machine inferencing over the ordered domain knowledge captured in OBO controlled vocabularies","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"1 1","pages":"145-154"},"PeriodicalIF":0.0,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84240833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper we propose new methods of chemical structure classification based on the integration of graph database mining from data mining and graph kernel functions from machine learning. In our method, we first identify a set of general graph patterns in chemical structure data. These patterns are then used to augment a graph kernel function that calculates the pairwise similarity between molecules. The obtained similarity matrix is used as input to classify chemical compounds via a kernel machines such as the support vector machine (SVM). Our results indicate that the use of a pattern-based approach to graph similarity yields performance profiles comparable to, and sometimes exceeding that of the existing state-of-the-art approaches. In addition, the identification of highly discriminative patterns for activity classification provides evidence that our methods can make generalizations about a compound's function given its chemical structure. While we evaluated our methods on molecular structures, these methods are designed to operate on general graph data and hence could easily be applied to other domains in bioinformatics.
{"title":"CHEMICAL COMPOUND CLASSIFICATION WITH AUTOMATICALLY MINED STRUCTURE PATTERNS.","authors":"A M Smalter, J Huan, G H Lushington","doi":"10.1901/jaba.2008.6-39","DOIUrl":"10.1901/jaba.2008.6-39","url":null,"abstract":"<p><p>In this paper we propose new methods of chemical structure classification based on the integration of graph database mining from data mining and graph kernel functions from machine learning. In our method, we first identify a set of general graph patterns in chemical structure data. These patterns are then used to augment a graph kernel function that calculates the pairwise similarity between molecules. The obtained similarity matrix is used as input to classify chemical compounds via a kernel machines such as the support vector machine (SVM). Our results indicate that the use of a pattern-based approach to graph similarity yields performance profiles comparable to, and sometimes exceeding that of the existing state-of-the-art approaches. In addition, the identification of highly discriminative patterns for activity classification provides evidence that our methods can make generalizations about a compound's function given its chemical structure. While we evaluated our methods on molecular structures, these methods are designed to operate on general graph data and hence could easily be applied to other domains in bioinformatics.</p>","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"6 ","pages":"39-48"},"PeriodicalIF":0.0,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2864492/pdf/nihms118197.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"28970890","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Proceedings of the 6th Asia-Pacific Bioinformatics Conference, APBC 2008, 14-17 January 2008, Kyoto, Japan","authors":"A. Brazma, S. Miyano, T. Akutsu","doi":"10.1142/p544","DOIUrl":"https://doi.org/10.1142/p544","url":null,"abstract":"","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"22 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82357635","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2007-12-01DOI: 10.1142/9781848161092_0009
Alper Küçükural, O. U. Sezerman, A. Erçil
Graph theoretic properties of proteins can be used to perceive the differences between correctly folded proteins and well designed decoy sets. 3D protein structures of proteins are represented with graphs. We used two different graph representations: Delaunay tessellations of proteins and contact map graphs. Graph theoretic properties for both graph types showed high classification accuracy for protein discrimination. Fisher, linear, quadratic, neural network, and support vector classifiers were used for the classification of the protein structures. The best classifier accuracy was over 98%. Results showed that characteristic features of graph theoretic properties can be used in the detection of native folds.
{"title":"Discrimination of Native Folds Using Network Properties of Protein Structures","authors":"Alper Küçükural, O. U. Sezerman, A. Erçil","doi":"10.1142/9781848161092_0009","DOIUrl":"https://doi.org/10.1142/9781848161092_0009","url":null,"abstract":"Graph theoretic properties of proteins can be used to perceive the differences between correctly folded proteins and well designed decoy sets. 3D protein structures of proteins are represented with graphs. We used two different graph representations: Delaunay tessellations of proteins and contact map graphs. Graph theoretic properties for both graph types showed high classification accuracy for protein discrimination. Fisher, linear, quadratic, neural network, and support vector classifiers were used for the classification of the protein structures. The best classifier accuracy was over 98%. Results showed that characteristic features of graph theoretic properties can be used in the detection of native folds.","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"291 1","pages":"59-68"},"PeriodicalIF":0.0,"publicationDate":"2007-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73729120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2007-12-01DOI: 10.1142/9781848161092_0027
Cuong V. Than, L. Nakhleh
The SPR (subtree prune and regraft) operation is used as the basis for reconciling incongruent phylogenetic trees, particularly for detecting and analyzing non-treelike evolutionary histories such as horizontal gene transfer, hybrid speciation, and recombination. The SPR-based tree reconciliation problem has been shown to be NP-hard, and several efficient heuristics have been designed to solve it. A major drawback of these heuristics is that for the most part they do not handle non-binary trees appropriately. Further, their computational efficiency suffers significantly when computing multiple optimal reconciliations. In this paper, we present algorithmic techniques for efficient SPR-based reconciliation of trees that are not necessarily binary. Further, we present divide-and-conquer approaches that enable efficient computing of multiple optimal reconciliations. We have implemented our techniques in the PhyloNet software package, which is publicly available at http://bioinfo.cs.rice.edu. The resulting method outperforms all existing methods in terms of speed, and performs at least as well as those methods in terms of accuracy.
{"title":"SPR-based Tree Reconciliation: Non-binary Trees and Multiple Solutions","authors":"Cuong V. Than, L. Nakhleh","doi":"10.1142/9781848161092_0027","DOIUrl":"https://doi.org/10.1142/9781848161092_0027","url":null,"abstract":"The SPR (subtree prune and regraft) operation is used as the basis for reconciling incongruent phylogenetic trees, particularly for detecting and analyzing non-treelike evolutionary histories such as horizontal gene transfer, hybrid speciation, and recombination. The SPR-based tree reconciliation problem has been shown to be NP-hard, and several efficient heuristics have been designed to solve it. A major drawback of these heuristics is that for the most part they do not handle non-binary trees appropriately. Further, their computational efficiency suffers significantly when computing multiple optimal reconciliations. In this paper, we present algorithmic techniques for efficient SPR-based reconciliation of trees that are not necessarily binary. Further, we present divide-and-conquer approaches that enable efficient computing of multiple optimal reconciliations. We have implemented our techniques in the PhyloNet software package, which is publicly available at http://bioinfo.cs.rice.edu. The resulting method outperforms all existing methods in terms of speed, and performs at least as well as those methods in terms of accuracy.","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"98 1","pages":"251-260"},"PeriodicalIF":0.0,"publicationDate":"2007-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83581096","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2007-12-01DOI: 10.1142/9781848161092_0014
H. Rangwala, G. Karypis
As the sequence identity between a pair of proteins decreases, alignment strategies that are based on sequence and/or sequence profiles become progressively less effective in identifying the correct structural correspondence between residue pairs. This significantly reduces the ability of comparative modelingbased approaches to build accurate structural models. Incorporating into the alignment process predicted information about the local structure of the protein holds the promise of significantly improving the alignment quality of distant proteins. This paper studies the impact on the alignment quality of a new class of predicted local structural features that measure how well fixed-length backbone fragments centered around each residue-pair align with each other. It presents a comprehensive experimental evaluation comparing these new features against existing state-of-the-art approaches utilizing profile-based and predicted secondary-structure information. It shows that for protein pairs with low sequence similarity (less than 12% sequence identity) the new structural features alone or in conjunction with profile-based information lead to alignments that are considerably better than those obtained by previous schemes.
{"title":"fRMSDAlign: Protein Sequence Alignment Using Predicted Local Structure Information for Pairs with Low Sequence Identity","authors":"H. Rangwala, G. Karypis","doi":"10.1142/9781848161092_0014","DOIUrl":"https://doi.org/10.1142/9781848161092_0014","url":null,"abstract":"As the sequence identity between a pair of proteins decreases, alignment strategies that are based on sequence and/or sequence profiles become progressively less effective in identifying the correct structural correspondence between residue pairs. This significantly reduces the ability of comparative modelingbased approaches to build accurate structural models. Incorporating into the alignment process predicted information about the local structure of the protein holds the promise of significantly improving the alignment quality of distant proteins. This paper studies the impact on the alignment quality of a new class of predicted local structural features that measure how well fixed-length backbone fragments centered around each residue-pair align with each other. It presents a comprehensive experimental evaluation comparing these new features against existing state-of-the-art approaches utilizing profile-based and predicted secondary-structure information. It shows that for protein pairs with low sequence similarity (less than 12% sequence identity) the new structural features alone or in conjunction with profile-based information lead to alignments that are considerably better than those obtained by previous schemes.","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"69 1","pages":"111-122"},"PeriodicalIF":0.0,"publicationDate":"2007-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82345935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2007-12-01DOI: 10.1142/9781848161092_0035
Francis Y. L. Chin, Henry C. M. Leung, M. Siu, S. Yiu
Finding motifs and the corresponding binding sites is a critical and challenging problem in studying the process of gene expression. String and matrix representations are two popular models to represent a motif. However, both representations share an important weakness by assuming that the occurrence of a nucleotide in a binding site is independent of other nucleotides. More complicated representations, such as HMM or regular expression, exist that can capture the nucleotide dependency. Unfortunately, these models are not practical (with too many parameters and require many known binding sites). Recently, Chin and h u n g introduced the SPSP representation which overcomes the limitations of these complicated models. However, discovering novel motifs in SPSP representation is still a NP-hard problem. In this paper, based on our observations in real binding sites, we propose a simpler model, the Dependency Pattern Sets (DPS) representation, which is simpler than the SPSP model but can still capture the nucleotide dependency. We develop a branch and bound algorithm (DPS-Finder) for finding optimal DPS motifs. Experimental results show that DPS-Finder can discover a length-10 motif from 22 length500 DNA sequences within a few minutes and the DPS representation has a similar performance as SPSP representation.
{"title":"Optimal Algorithm for Finding DNA Motifs with Nucleotide Adjacent Dependency","authors":"Francis Y. L. Chin, Henry C. M. Leung, M. Siu, S. Yiu","doi":"10.1142/9781848161092_0035","DOIUrl":"https://doi.org/10.1142/9781848161092_0035","url":null,"abstract":"Finding motifs and the corresponding binding sites is a critical and challenging problem in studying the process of gene expression. String and matrix representations are two popular models to represent a motif. However, both representations share an important weakness by assuming that the occurrence of a nucleotide in a binding site is independent of other nucleotides. More complicated representations, such as HMM or regular expression, exist that can capture the nucleotide dependency. Unfortunately, these models are not practical (with too many parameters and require many known binding sites). Recently, Chin and h u n g introduced the SPSP representation which overcomes the limitations of these complicated models. However, discovering novel motifs in SPSP representation is still a NP-hard problem. In this paper, based on our observations in real binding sites, we propose a simpler model, the Dependency Pattern Sets (DPS) representation, which is simpler than the SPSP model but can still capture the nucleotide dependency. We develop a branch and bound algorithm (DPS-Finder) for finding optimal DPS motifs. Experimental results show that DPS-Finder can discover a length-10 motif from 22 length500 DNA sequences within a few minutes and the DPS representation has a similar performance as SPSP representation.","PeriodicalId":74513,"journal":{"name":"Proceedings of the ... Asia-Pacific bioinformatics conference","volume":"24 1","pages":"343-352"},"PeriodicalIF":0.0,"publicationDate":"2007-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83220207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号