MICAlign: a sequence-to-structure alignment tool integrating multiple sources of information in conditional random fields. Summary: Sequence-to-structure alignment in template-based protein structure modeling for remote homologs remains a difficult problem even following the correct recognition of folds. Here we present MICAlign, a sequence-to-structure alignment tool that incorporates multiple sources of information from local structural contexts of template, sequence profiles, predicted secondary structures, solvent accessibilities, potential-like terms (including residue–residue contacts and solvent exposures) and pre-aligned structures and sequences. These features, together with a position-specific gap scheme, were integrated into conditional random fields through which the optimal parameters were automatically learned. MICAlign showed improved alignment accuracy over several other state-of-the-art alignment tools based on comparisons by using independent datasets. Availability: Freely available at http://www.bioinfo.tsinghua.edu.cn/∼xiaxf/micalign for both web server and source code.
References in zbMATH (referenced in 2 articles )
Showing results 1 to 2 of 2.
- Daugelaite, Jurate; O’Driscoll, Aisling; Sleator, Roy D.: An overview of multiple sequence alignments and cloud computing in bioinformatics (2013)
- Xia, Xuefeng; Zhang, Song; Su, Yu; Sun, Zhirong: Micalign: a sequence-to-structure alignment tool integrating multiple sources of information in conditional random fields (2009)