Prioritizing functional phosphorylation sites based on multiple feature integration

Abstract

Protein phosphorylation is an important type of post-translational modification that is involved in a variety of biological activities. Most phosphorylation events occur on serine, threonine and tyrosine residues in eukaryotes. In recent years, many phosphorylation sites have been identified as a result of advances in mass-spectrometric techniques. However, a large percentage of phosphorylation sites may be non-functional. Systematically prioritizing functional sites from a large number of phosphorylation sites will be increasingly important for the study of their biological roles. This study focused on exploring the intrinsic features of functional phosphorylation sites to predict whether a phosphosite is likely to be functional. We found significant differences in the distribution of evolutionary conservation, kinase association, disorder score, and secondary structure between known functional and background phosphorylation datasets. We built four different types of classifiers based on the most representative features and found that their performances were similar. We also prioritized 213,837 human phosphorylation sites from a variety of phosphorylation databases, which will be helpful for subsequent functional studies. All predicted results are available for query and download on our website (Predict Functional Phosphosites, PFP, http://pfp.biosino.org/).

Proteome-wide Structural Analysis of PTM Hotspots Reveals Regulatory Elements Predicted to Impact Biological Function and Disease

“…S4). During the review of this manuscript, we discovered that an alternative modeling approach was taken in which some (but not all) similar features were used to generate predictive models for phosphorylation (47). Although this suggests that there are indeed more features to be considered, neither the distribution of known function source count nor the contributions of individual features were fully analyzed.…”

SAPH-ire TFx – A Recommendation-based Machine Learning Model Captures a Broad Feature Landscape Underlying Functional Post-Translational Modifications

“…PTM site features such as evolutionary conservation (14–16), PTM co-localization (17–19), protein structural constraints (20,21), and other single features, have been shown to be mildly predictive for function and offer useful means of filtering PTM data for likely important modifications. More recently, machine learning models that incorporate multiple single features have been proven to have significantly greater predictive power in comparison to single features alone (22–24). In general, these models provide a priority score that is effective for rank ordering PTM function potential, and are shown to be successful in identifying new functional PTMs across the eukaryotic proteome, but unfortunately do little to confidently recommend PTMs for experimental analysis due to very poor recall of true positive data.…”

Scop3P: a comprehensive resource of human phosphosites within their full context

“…Phosphatases, the erasers of the phosphate group, function opposite to kinases: they remove the phosphate group from phosphorylated residues 7,8 . Several studies have attempted to differentiate between functional and non-functional P-sites based on their evolutionary conservation 9,10 , kinase specificity 11,12 , PTM cross-talk or based on their interactions 13 . However, P-site conservation may not be particularly useful to determine functional importance of a P-site as only a small fraction (~35%) of functional P-sites were reported to be conserved 14,15 , while some functional P-sites have been identified in poorly conserved regions 11,16 .…”