A comprehensive comparative review of sequencebased predictors of DNA-and RNA-binding residues.

Motivated by the pressing need to characterize protein-DNA and protein-RNAinteractions on large scale, we review a comprehensive set of 30 computational methods for high-throughput prediction of RNA-or DNA-binding residues fromprotein sequences. Wesummarize these predictors fromseveral significant perspectives including their design, outputs and availability. Weperformempirical assessment ofmethods that offer web servers using a new benchmark data set characterized by a more complete annotation that includes binding residues transferred fromthe same or similar proteins.Weshowthat predictors of DNA-binding (RNA-binding) residues offer relatively strong predictive performance but they are unable to properly separate DNA- fromRNA-binding residues.Wedesign andempirically assess several types of consensuses and demonstrate that machine learning (ML)-based approaches provide improved predictive performancewhen compared with the individual predictors of DNA-binding residues or RNA-binding residues.Wealso formulate and execute first-of-its-kind study that targets combined prediction ofDNA- and RNA-binding residues.Wedesign and test three types of consensuses for this prediction and conclude that this novel approach that relies on ML design provides better predictive quality than individual predictors when tested on prediction ofDNA- and RNA-binding residues individually. It also substantially improves discrimination between these two types of nucleic acids. Our results suggest that development of a new generation of predictors would benefit fromusing training data sets that combine both RNA- andDNA-binding proteins, designing new inputs that specifically target either DNA- or RNA-binding residues and pursuing combined prediction of DNA- and RNA-binding residues. [ABSTRACT FROM AUTHOR]