Towards the prediction of essential genes by integration of network topology, cellular localization and biological process information

Abstract

Background: The identification of essential genes is important for the understanding of the minimal requirements for cellular life and for practical purposes, such as drug design. However, the experimental techniques for essential genes discovery are labor-intensive and time-consuming. Considering these experimental constraints, a computational approach capable of accurately predicting essential genes would be of great value. We therefore present here a machine learningbased computational approach relying on network topological features, cellular localization and biological process information for prediction of essential genes.

Topological properties of robust biological and computational networks

“…Anecdotal evidence suggests that modulelevel properties may be important in determining network robustness to mutations. For example, cellular localization was found to be a useful feature when predicting gene essentiality in yeast [18], and nuclear proteins in particular were shown to be enriched for such essential genes [19]. Other studies have also shown that involvement of genes in specific subnetworks is the most indicative of individual gene essentiality [20,21].…”

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

“…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).…”

“…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

“…Besides, some intrinsic complications exist in the experimental ways. When considering much more uncultured microorganisms (Chitsaz et al, 2011), the experimental methods for essential gene identification would further show its limitation (Acenico and Lemke, 2009;Holman et al, 2009). Considering the importance of essential genes and the limitation of experimental methods, predicting essential genes in silico is of paramount importance.…”

“…Using only homology information, some species-specific genes may be classified into the wrong class (Deng et al, 2011b), and it takes a long time to search for the homology. Combining all of the features mentioned above to predict essential genes was also an important strategy for some methods (Gustafson et al, 2006;Roberts et al, 2007;Acenico and Lemke, 2009;Deng et al, 2011a;Lin and Zhang, 2011). If a query essential gene has no homologs in public databases and no prior experimental information, it is impossible to predict its essentiality.…”