ComSin: database of protein structures in bound (complex) and unbound (single) states in relation to their intrinsic disorder

Abstract

Most of the proteins in a cell assemble into complexes to carry out their function. In this work, we have created a new database (named ComSin) of protein structures in bound (complex) and unbound (single) states to provide a researcher with exhaustive information on structures of the same or homologous proteins in bound and unbound states. From the complete Protein Data Bank (PDB), we selected 24 910 pairs of protein structures in bound and unbound states, and identified regions of intrinsic disorder. For 2448 pairs, the proteins in bound and unbound states are identical, while 7129 pairs have sequence identity 90% or larger. The developed server enables one to search for proteins in bound and unbound states with several options including sequence similarity between the corresponding proteins in bound and unbound states, and validation of interaction interfaces of protein complexes. Besides that, through our web server, one can obtain necessary information for studying disorder-to-order and order-to-disorder transitions upon complex formation, and analyze structural differences between proteins in bound and unbound states. The database is available at http://antares.protres.ru/comsin/.

Figure 1.

ADP-ribosylation factor (green and blue chains) and exchange factor ARNO (magenta and yellow chains) in unbound (left; PDB entries 1RRF and 1R8M, correspondingly) and bound (right; PDB entry 1R8Q) forms. Blue, red and orange dashed lines correspond to disordered regions at N-terminus, at C-terminus and in the central part of protein chains, correspondingly.

Figure 2.

Dependence of the number of complex–single pairs on the sequence identity cutoff. There are 2448, 6051 and 7129 single–complex pairs at 100, 95 and 90% identity level cutoffs, respectively.

Figure 3.

A histogram of the number of proteins in the database by the number of disordered residues in the structure, for 2448 pairs of homologous proteins with 100% identity level.

Figure 4.

(A) A screenshot of ComSin results filtered for sequence identity 100%. (B) A screenshot of an individual page with information on disordered regions for PDB entry 2D6K.