Predicting gene–phenotype associations in humans and other species from orthologous and paralogous phenotypes
MetadataShow full item record
Phenotypes and diseases may be related by seemingly dissimilar phenotypes in other species by means of the orthology of underlying genes. Such "orthologous phenotypes," or "phenologs," are examples of deep homology, and one member of the orthology relationship may be used to predict candidate genes for its counterpart. (There exists evidence of "paralogous phenotypes" as well, but validation is non-trivial.) In Chapter 2, I demonstrate the utility of including plant phenotypes in our database, and provide as an example the prediction of mammalian neural crest defects from an Arabidopsis thaliana phenotype, negative gravitropism defective. In the third chapter, I describe the incorporation of additional phenotypes into our database (including chicken, zebrafish, E. coli, and new C. elegans datasets). I present a method, developed in coordination with Martin Singh-Blom, for ranking predicted candidate genes by way of a k nearest neighbors naïve Bayes classifier drawing phenolog information from a variety of species. The fourth chapter relates to a computational method and application for identifying shared and overlapping pathways which contribute to phenotypes. I describe a method for rapidly querying a database of phenotype--gene associations for Boolean combinations of phenotypes which yields improved predictions. This method offers insight into the divergence of orthologous pathways in evolution. I demonstrate connections between breast cancer and zebrafish methylmercury response (through oxidative stress and apoptosis); human myopathy and plant red light response genes, minus those involved in water deprivation response (via autophagy); and holoprosencephaly and an array of zebrafish phenotypes. In the first appendix, I present the SciRuby Project, which I co-founded in order to bring scientific libraries to the Ruby programming language. I describe the motivation behind SciRuby and my role in its creation. Finally in Appendix B, I discuss the first beta release of NMatrix, a dense and sparse matrix library for the Ruby language, which I developed in part to facilitate and validate rapid phenolog searches. In this work, I describe the concept of phenologs as well as the development of the necessary computational tools for discovering phenotype orthology relationships, for predicting associated genes, and for statistically validating the discovered relationships and predicted associations.
Showing items related by title, author, creator and subject.
McGary, Kriston L.; Lee, Insuk; Marcotte, Edward M. (Genome Biology, 2007-12-05)We demonstrate that loss-of-function yeast phenotypes are predictable by guilt-by-association in functional gene networks. Testing 1,102 loss-of-function phenotypes from genome-wide assays of yeast reveals predictability ...
Woods, John O.; Singh-Blom, Ulf Martin; Laurent, Jon M.; McGary, Kriston L.; Marcotte, Edward M. (BMC Bioinformatics, 2013-06-21)Background: Phenotypes and diseases may be related to seemingly dissimilar phenotypes in other species by means of the orthology of underlying genes. Such “orthologous phenotypes,” or “phenologs,” are examples of deep ...
Evolutionary genetics and ecology of water use efficiency ([delta]¹³C) in Ipomopsis agregata and Arabidopsis thaliana Kenney, Amanda Marie (2011-12)My dissertation research investigates the genetic architecture and evolutionary significance of physiological variation in two wildflower species, Ipomopsis aggregata and Arabidopsis thaliana. In particular, my work focuses ...