Open AccessCompleteness and Consistency in Structural Domain Classifications
Author(s) -
R. Dustin Schaeffer,
Lisa N. Kinch,
Jimin Pei,
Kirill E. Medvedev,
Nick V. Grishin
Publication year - 2021
Publication title -
acs omega
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.779
H-Index - 40
ISSN - 2470-1343
DOI - 10.1021/acsomega.1c00950
Subject(s) - domain (mathematical analysis) , major facilitator superfamily , computer science , protein data bank (rcsb pdb) , scope (computer science) , pairwise comparison , completeness (order theory) , consistency (knowledge bases) , computational biology , data mining , artificial intelligence , biology , mathematics , genetics , biochemistry , programming language , mathematical analysis , mutant , gene
Domain classifications are a useful resource for computational analysis of the protein structure, but elements of their composition are often opaque to potential users. We perform a comparative analysis of our classification ECOD against the SCOPe, SCOP2, and CATH domain classifications with respect to their constituent domain boundaries and hierarchal organization. The coverage of these domain classifications with respect to ECOD and to the PDB was assessed by structure and by sequence. We also conducted domain pair analysis to determine broad differences in hierarchy between domains shared by ECOD and other classifications. Finally, we present domains from the major facilitator superfamily (MFS) of transporter proteins and provide evidence that supports their split into domains and for multiple conformations within these families. We find that the ECOD and CATH provide the most extensive structural coverage of the PDB. ECOD and SCOPe have the most consistent domain boundary conditions, whereas CATH and SCOP2 both differ significantly.