Allows users to predict heme binding residues (HEMEs). For HEMEsPred, several sequences and structure-based features (such as amino acid composition) were collected to construct feature matrices. It contains an adaptive ensemble learning scheme designed to treat of the class-imbalance problem as well as to enhance the prediction performance. Moreover, different ligand-specific models are also included, and they consider that different heme ligands can be varied in their roles, sizes and distributions.
School of Computer Science and Information Technology, Northeast Normal University, Changchun, China; Department of Computer Science, School of Engineering, Virginia Commonwealth University, Richmond, VA, USA
HEMEsPred funding source(s)
Supported by the Fundamental Research Funds for the Central Universities (Grant No. 14ZZ2240), and the China Scholarship Council.