Abstrakt

Integrative support vector machine for the prediction of zinc-binding sites in proteins

Hui Li, Dechang Pi, Lihang Zhang, Lei Hong

Zinc binding proteins play an important role in biological function, many researches focus on the area of zinc-binding sites. Taking into account the advantages of support vector machine, based on the different tools for the prediction of zinc-binding sites, a novel predictor named combZincPred was proposed to integrate these result scores. Tested on a non-redundant dataset, AURPC of our method increased more, and other indexes are also better than the other three predictors. The method can be better used to the inference of zinc-binding protein function.