Automated Gene Function Prediction

The objective of biological research is to understand the structural and the functional aspects of life. Though living organisms are diverse in almost every aspect, they are made of cells, and share the same machinery for their basic functions. The structural and functional aspect of life is traceable to genes, given that the information from the genes determine the protein composition and thereby the function of the cell. Hence, predicting the functions of individual genes is the gate way for understanding the blueprint of life. The rationale behind the ongoing genome sequencing projects is to utilize the sequence information to understand the genes and their functions. The exponential increase in the amount of sequence information enunciated the need for an automated approach to acquire knowledge about their biological function. This book introduces the general strategies used in the automated annotation of genes and protein sequences. Specifically, it describes a method utilizing the machine learning approach to predict gene function. This book is addressed to researchers involved in predicting gene function and applying machine learning algorithms to other biological problems.

This volume introduces software used for gene prediction with focus on eukaryotic genomes. The chapters in this book describe software and web server usage as applied in common use-cases, and explain ways to simplify re-annotation of long available genome assemblies. Written in the highly successful Methods in Molecular Biology series format, chapters include introductions to their respective topics, lists of the necessary computational requirements, step-by-step, readily reproducible computational protocols, and tips on troubleshooting and avoiding known pitfalls. Cutting-edge and thorough, Gene Prediction: Methods and Protocols is a valuable resource for researchers and research groups working on the assembly and annotation of single species or small groups of species. Chapter 3 is available open access under a CC BY 4.0 license via link.springer.com.

This book provides a practical and self-contained overview of the Gene Ontology (GO), the leading project to organize biological knowledge on genes and their products across genomic resources. Written for biologists and bioinformaticians, it covers the state-of-the-art of how GO annotations are made, how they are evaluated, and what sort of analyses can and cannot be done with the GO. In the spirit of the Methods in Molecular Biology book series, there is an emphasis throughout the chapters on providing practical guidance and troubleshooting advice. Authoritative and accessible, The Gene Ontology Handbook serves non-experts as well as seasoned GO users as a thorough guide to this powerful knowledge system. This work was published by Saint Philip Street Press pursuant to a Creative Commons license permitting commercial use. All rights not granted by the work's license are retained by the author or authors.

Numerous genome projects have produced a large and ever increasing amount of genomic sequence data. However, the biological functions of many proteins encoded by the sequences remain unknown. Protein function annotation and prediction become an essential and challenging task of post-genomic research. In this research, we present an automated protein function prediction system based on a set of proteins of known biological functions. The functions of the proteins are characterized with Gene Ontology (GO) annotations. The prediction system uses a novel measure to calculate the pair-wise overall similarity between protein sequences. The protein function prediction is performed based on the GO annotations of similar sequences using a weighted k-nearest neighbor method. We show the prediction accuracies obtained using the model organism yeast (Sacchyromyces cerevisiae). The results indicate that the weighted k-nearest neighbor method significantly outperforms the regular k-nearest neighbor method for protein biological function prediction.

Within the last decade, much progress has been made in the analysis and diagnosis of human inherited disease, and in the characterization of the underlying genes and their associated pathological lesions.

Sequence - Evolution - Function is an introduction to the computational approaches that play a critical role in the emerging new branch of biology known as functional genomics. The book provides the reader with an understanding of the principles and approaches of functional genomics and of the potential and limitations of computational and experimental approaches to genome analysis. Sequence - Evolution - Function should help bridge the "digital divide" between biologists and computer scientists, allowing biologists to better grasp the peculiarities of the emerging field of Genome Biology and to learn how to benefit from the enormous amount of sequence data available in the public databases. The book is non-technical with respect to the computer methods for genome analysis and discusses these methods from the user's viewpoint, without addressing mathematical and algorithmic details. Prior practical familiarity with the basic methods for sequence analysis is a major advantage, but a reader without such experience will be able to use the book as an introduction to these methods. This book is perfect for introductory level courses in computational methods for comparative and functional genomics.

The human genome project and numerous other genome projects have produced a large and ever increasing amount of sequence data. One of the main research challenges in the post-genomic era is to understand the relationship between the nucleotide sequences of genes and the functions of the proteins they encode. The objective of this thesis is to develop an automated protein function prediction system that is based on a set of homologous proteins and gene ontology categories. A novel measure based on a set of best local alignments is used to identify the homologues. The biological functions of the homologous proteins are characterized with gene ontology annotations. The protein function prediction is performed based on data mining models using decision trees. The models are trained and tested using the complete proteome of model organism yeast. The results show that the prediction accuracy depends on individual functional groups of proteins. There is a general trend of decreased model accuracy with the level of a group on the gene ontology graph, but the accuracy at a fix level varies from group to group. The prediction accuracy varies from group to group, no obvious accuracy changes from one level to another. These variations of accuracy illustrate certain limitations of sequence-based protein function prediction methods. But the fundamental assumption used in this thesis, similar amino acid sequences implying similar biological functions, is largely valid. The prediction results based on the proteome of yeast indicate that the accuracies for most of the functional groups are over 75%. We conclude that the decision tree model can be used as a preliminary tool for protein function prediction although the prediction results need to be verified through other means.

This contributed volume explores the emerging intersection between big data analytics and genomics. Recent sequencing technologies have enabled high-throughput sequencing data generation for genomics resulting in several international projects which have led to massive genomic data accumulation at an unprecedented pace. To reveal novel genomic insights from this data within a reasonable time frame, traditional data analysis methods may not be sufficient or scalable, forcing the need for big data analytics to be developed for genomics. The computational methods addressed in the book are intended to tackle crucial biological questions using big data, and are appropriate for either newcomers or veterans in the field.This volume offers thirteen peer-reviewed contributions, written by international leading experts from different regions, representing Argentina, Brazil, China, France, Germany, Hong Kong, India, Japan, Spain, and the USA. In particular, the book surveys three main areas: statistical analytics, computational analytics, and cancer genome analytics. Sample topics covered include: statistical methods for integrative analysis of genomic data, computation methods for protein function prediction, and perspectives on machine learning techniques in big data mining of cancer. Self-contained and suitable for graduate students, this book is also designed for bioinformaticians, computational biologists, and researchers in communities ranging from genomics, big data, molecular genetics, data mining, biostatistics, biomedical science, cancer research, medical research, and biology to machine learning and computer science. Readers will find this volume to be an essential read for appreciating the role of big data in genomics, making this an invaluable resource for stimulating further research on the topic.