Protein disorderness based prediction of essential genes of Saccharomyces cerevisiae: A machine learning approach

Protein structure is phyletically conserved to serve its specific functions. But several proteins are found to be partially lacking a definite folded structure under specific conditions and this is known as disorderness. Intrinsically disordered proteins serve many important functions of the cell. Essential genes of an organism refer to the minimal gene set among the genome and mutation in any of which may confer a lethal or non-fertile phenotype. Essential genes show high phyletic retention and high degree of conservation. Thus comparison on the basis of disorderness of the proteins encoded by essential and non-essential genes may form a basis of segregation between these two kinds of genes. A machine learning framework using neural network as a classifier was implemented here which could successfully segregate the essential genes from non-essential ones.