Predicting essential genes in fungal genomes

Abstract

Essential genes are required for an organism's viability, and the ability to identify these genes in pathogens is crucial to directed drug development. Predicting essential genes through computational methods is appealing because it circumvents expensive and difficult experimental screens. Most such prediction is based on homology mapping to experimentally verified essential genes in model organisms. We present here a different approach, one that relies exclusively on sequence features of a gene to estimate essentiality and offers a promising way to identify essential genes in unstudied or uncultured organisms. We identified 14 characteristic sequence features potentially associated with essentiality, such as localization signals, codon adaptation, GC content, and overall hydrophobicity. Using the well-characterized baker's yeast Saccharomyces cerevisiae, we employed a simple Bayesian framework to measure the correlation of each of these features with essentiality. We then employed the 14 features to learn the parameters of a machine learning classifier capable of predicting essential genes. We trained our classifier on known essential genes in S. cerevisiae and applied it to the closely related and relatively unstudied yeast Saccharomyces mikatae. We assessed predictive success in two ways: First, we compared all of our predictions with those generated by homology mapping between these two species. Second, we verified a subset of our predictions with eight in vivo knockouts in S. mikatae, and we present here the first experimentally confirmed essential genes in this species.

Characteristics of Plant Essential Genes Allow for within- and between-Species Prediction of Lethal Mutant Phenotypes

“…Similar to findings in yeast and mouse (Seringhaus et al, 2006;Yuan et al, 2012), these features can be used to predict genes with lethal phenotypes in plants. We also show that lethal-phenotype gene prediction models can be applied across species with reasonable performance.…”

“…To address these questions, we applied machine learning methods that have been used for essential gene predictions in budding yeast (Seringhaus et al, 2006;Acencio and Lemke, 2009) and mouse (Yuan et al, 2012). A matrix of genes with a documented phenotype and their associated values for different features (Supplemental Data Set 3) was used as input for six machine learning classifiers (see Methods; Figure 6A).…”

“…Considering that some of the features we found to be correlated with gene lethality have been shown to be important in other species (Seringhaus et al, 2006;Yuan et al, 2012), this raises the question whether a prediction model trained with A. thaliana data (A. thaliana model) can be used to predict phenotype lethality across species boundaries. To test this, we first collected rice phenotype data for 92 genes (18 lethal; see Methods; Supplemental Data Set 1) and analogous genomics and functional annotation data (Table 1).…”

“…Some of these attributes are shared by lethal-phenotype genes in C. elegans, S. pombe, and Mus musculus (Kamath et al, 2003;Kim et al, 2010;Yuan et al, 2012). Using these features, lethal-phenotype genes have been predicted in S. cerevisiae and M. musculus (Seringhaus et al, 2006;Acencio and Lemke, 2009;Yuan et al, 2012).…”

Predicting bacterial essential genes using only sequence composition information

“…In 2006, Seringhaus et al (2006) combined 7 different machine learning methods to predict essential genes in fungal genomes using 14 sequence compositional features. Some of the predicted essential genes were chosen and validated by experiments.…”