DOMAC is an accurate, hybrid protein domain prediction server. DOMAC integrates homology modeling, domain parsing, and ab intio methods together. The preliminary implementation of DOMAC (server name: FOLDpro (Cheng and Baldi, Bioinformatics, 2006) ) is ranked first among all domain prediction servers in the seventh edition of Critical Assessment of Techniques for Protein Structure Prediction (CASP7) (Moult et al., Proteins, 2005). DOMAC predicts protein domains in the following two steps.
First, it uses PSI-BLAST (Altschul et al., NAR, 1997) to search the target sequence against NCBI Non-Redundant sequence database to construct a profile. The profile is used to search a template library built from the proteins in Protein Data Bank (Berman et al., Nucleic Acids Res., 2000) to identify templates.
Second, if some significant templates are identified (e-value <= 0.001), it generates a structure model for the target using Modeller (Sali and Blundell, JMB, 1993) based on the template structures. Multiple significant templates are combined to improve model quality if available. Then it uses a domain parsing tool PDP ((Alexandrov and Shindyalov, Bioinformatics, 2003) to parse the model into domains. If the parsed domains do not cover the whole target sequence, DOMAC will assign the uncovered regions to the adjacent domains.
If no significant template is found, DOMAC will invoke the ab initio domain predictor DOMpro to predict domains. DOMpro (Cheng et al., Data Mining and Knowledge Discovery, 2006) uses neural networks in conjunction with sequence profile, predicted secondary structure, and relative solvent accessibility to predict domain boundary.