CTSS: A Robust and Efficient Method for Protein Structure Alignment Based on Local Geometrical and Biological Features

  • Authors:
  • Tolga Can;Yuan-Fang Wang

  • Affiliations:
  • -;-

  • Venue:
  • CSB '03 Proceedings of the IEEE Computer Society Conference on Bioinformatics
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present a new method for conducting proteinstructure similarity searches, which improves on theaccuracy, robustness, and efficiency of some existingtechniques. Our method is grounded in the theory ofdifferential geometry on 3D space curve matching. Wegenerate shape signatures for proteins that areinvariant, localized, robust, compact, and biologicallymeaningful. To improve matching accuracy, we smooththe noisy raw atomic coordinate data with spline fitting.To improve matching efficiency, we adopt a hierarchicalcoarse-to-fine strategy. We use an efficient hashing-basedtechnique to screen out unlikely candidates andperform detailed pairwise alignments only for a smallnumber of candidates that survive the screening process.Contrary to other hashing based techniques, ourtechnique employs domain specific information (not justgeometric information) in constructing the hash key, andhence, is more tuned to the domain of biology.Furthermore, the invariancy, localization, andcompactness of the shape signatures allow us to utilize awell-known local sequence alignment algorithm foraligning two protein structures. One measure of theefficacy of the proposed technique is that we were ableto discover new, meaningful motifs that were notreported by other structure alignment methods.