Predicting Gene Ontology Biological Process From Temporal Gene Expression Patterns

Abstract

The aim of the present study was to generate hypotheses on the involvement of uncharacterized genes in biological processes. To this end, supervised learning was used to analyze microarray-derived time-series gene expression data. Our method was objectively evaluated on known genes using cross-validation and provided high-precision Gene Ontology biological process classifications for 211 of the 213 uncharacterized genes in the data set used. In addition, new roles in biological process were hypothesized for known genes. Our method uses biological knowledge expressed by Gene Ontology and generates a rule model associating this knowledge with minimal characteristic features of temporal gene expression profiles. This model allows learning and classification of multiple biological process roles for each gene and can predict participation of genes in a biological process even though the genes of this class exhibit a wide variety of gene expression profiles including inverse coregulation. A considerable number of the hypothesized new roles for known genes were confirmed by literature search. In addition, many biological process roles hypothesized for uncharacterized genes were found to agree with assumptions based on homology information. To our knowledge, a gene classifier of similar scope and functionality has not been reported earlier.

Temporal Transcription Program of Recombinant Autographa californica Multiple Nucleopolyhedrosis Virus

“…Genes participating in the same biological process may be transcriptionally coregulated (2,27,38,54). Thus, baculovirus genes having the same temporal expression pattern during the time course of infection may imply sets of functionally correlated genes or suggest a common regulatory sequence motif in the genome.…”

Discovering regulatory binding-site modules using rule-based learning

“…However, we also see several other extensions within the framework of the present work leading to better understanding of the regulatory mechanisms. Since some transcription factors are only active at certain times or under certain conditions, a more advanced definition of coexpression which, for instance, takes into account correlation over subsets of expression time points might be advantageous (see Laegreid et al 2003). Another current research issue involves repeating the grouping of expression profiles and rule induction in a feedback loop.…”

A hierarchical multi-label classification ant colony algorithm for protein function prediction

“…Much work on hierarchical classification of protein functions has been focused on training a classifier for each class label (function) independently, using the hierarchy to determine positive and negative examples associated with each classifier [2,3,20,22]. As discussed in [6], predicting each class label individually has several disadvantages, as follows.…”