mTM-align: a server for fast protein structure database search and multiple protein structure alignment

Abstract

With the rapid increase of the number of protein structures in the Protein Data Bank, it becomes urgent to develop algorithms for efficient protein structure comparisons. In this article, we present the mTM-align server, which consists of two closely related modules: one for structure database search and the other for multiple structure alignment. The database search is speeded up based on a heuristic algorithm and a hierarchical organization of the structures in the database. The multiple structure alignment is performed using the recently developed algorithm mTM-align. Benchmark tests demonstrate that our algorithms outperform other peering methods for both modules, in terms of speed and accuracy. One of the unique features for the server is the interplay between database search and multiple structure alignment. The server provides service not only for performing fast database search, but also for making accurate multiple structure alignment with the structures found by the search. For the database search, it takes about 2–5 min for a structure of a medium size (∼300 residues). For the multiple structure alignment, it takes a few seconds for ∼10 structures of medium sizes. The server is freely available at: http://yanglab.nankai.edu.cn/mTM-align/.

RUPEE: A fast and accurate purely geometric protein structure search

“…32 We first discuss some related work to provide a context for our approach followed by 33 a description of our method. We end with a comparison of results against the 34 mTM-align structure search [5], the secondary structure matching (SSM) search [6], and 35 the CATHEDRAL structural scan [7] available at the CATH website. 36 Related work 37 Pairwise alignment involves finding a set of spatial rotations and translations for two 38 protein structures that minimizes a distance metric.…”

“…As an example of our linear encoding, we apply our method to the β-turn-β motif shown in Fig 3. The corresponding sequence of residue descriptors is shown below. [ 5,5,5,5,5,5,7,5,11,11,5,5,5,5,5,5 ] (1)…”

“…{ [5,5,5], [5,5,5], [5,5,5], [5,5,5], [5,5,7] [5, 7, 5], [7,5,11], [5,11,11], [11,11,5], [11,5,5] [5, 5, 5], [5,5,5], [5,5,5], [5,5,5]…”

“…221 For our results, we have created 3 benchmarks, scop d360, scop d62, and cath d99, 222 for pairwise evaluations to mTM, SSM, and CATHEDRAL, respectively. scop d360 is 223 derived from the d500 benchmark used in [5] filtered for domains in SCOPe 2.07 for 224 which mTM returns 100 or more results. Similarly, scop d62 is derived from the d500 225 benchmark filtered for domains defined in SCOP 1.73 for which SSM returns 50 or more 226 results.…”

“…becomes the following bag of integers.Notice in (4) the value 915, corresponding to the shingle[5,5,5], occurs frequently 144 indicating the presence of β-strands. Since most proteins are dominated by regular 145 secondary structure, the abundance of shingles for β-strands as well as the three types 146 of helices, end up dominating comparisons.…”