Accurate in silico identification of species-specific acetylation sites by integrating protein sequence-derived and functional features

摘要: Lysine acetylation is a reversible post-translational modification, playing an important role in cytokine signaling, transcriptional regulation, and apoptosis. To fully understand mechanisms, identification of substrates specific sites crucial. Experimental often time-consuming expensive. Alternative bioinformatics methods are cost-effective can be used high-throughput manner to generate relatively precise predictions. Here we develop method termed as SSPKA for species-specific lysine prediction, using random forest classifiers that combine sequence-derived functional features with two-step feature selection. Feature importance analysis indicates features, applied site prediction the first time, significantly improve predictive performance. We apply model screen entire human proteome identify many high-confidence putative not previously identified. The results along implemented Java tool, serve useful resources elucidate mechanism facilitate hypothesis-driven experimental design validation.