Revista de minería de datos en genómica y proteómica

Abstracto

Units: Universal True SDSA (Structure-Dependent Sequence Alignment)

Scott Foy and Gerald Wyckoff

Motivation: The Universal True SDSA (Structure-dependent Sequence Alignment), or UniTS, program calculates the most probable amino acid sequence alignment derived from multiple superimposed protein three-dimensional structures. Additionally, utilizing this newly generated SDSA, UniTS calculates improved quality assessment scores (e.g., RMSD, etc.) for the superimposed protein structures. Although other algorithms have been developed to derive an amino acid sequence alignment from aligned protein three-dimensional structures utilizing atomic proximity, none of these appropriately manages multiple residue matches, prevents the incorrect ordering of residues, and sequentially aligns structurally nonconserved regions. UniTS compensates for the weaknesses inherent in residue profile-based SDSA programs and structural alignment programs. Unlike the residue profile-based SDSA programs utilized as precursors to threading and homology modeling, UniTS is truly structure-dependent. Results: The results presented herein demonstrate that UniTS calculates the universal sequence alignment for the complete protein compared to the partial sequence alignment derived from structural alignment programs. Furthermore, these results demonstrate the capability of UniTS to refine the sequence alignment input into a superpositioning program and utilize this refined alignment to calculate improved structural quality assessment scores. Finally, the quality score generation capabili