Date |
Jul 21, 2017 |
Speaker |
Dr. Kentaro Tomii,
Intelligent Bioinformatics Research Team, Artificial Intelligence Research Center, Advanced Industrial Science and Technology (AIST)
|
Title |
A novel amino acid substitution matrix for identifying distantly related proteins
|
Abstract |
There have been reports of various amino acid substitution matrices to search for sequence similarity of proteins. Recently, we have developed a novel sensitive matrix, called MIQS, which was generated from the principal component analysis (PCA) subspace based on the three series of typical existing matrices by combining benchmark results and kernel density estimation (KDE). This novel matrix shows better detection performance for identifying distantly related proteins than those of existing matrices. Some tools, such as DECIPHER, LAST, and FAMSA adopted this matrix. MIQS is available at http://csas.cbrc.jp/Ssearch/ .
References
Yamada et al., “Revisiting amino acid substitution matrices for identifying distantly related proteins.” Bioinformatics 30: 317?325 (2014). doi:10.1093/bioinformatics/btt694
Tomii et al., “Systematic Exploration of an Efficient Amino Acid Substitution Matrix: MIQS.” Methods Mol. Biol. 1415: 211?223 (2016). doi:10.1007/978-1-4939-3572-7_11
Lim et al., “Protein sequence-similarity search acceleration using a heuristic algorithm with a sensitive matrix.” J. Struct. Funct. Genomics 17: 147-154 (2016). doi: 10.1007/s10969-016-9210-4
|
|