文档介绍：A Segment Matching Method for Remote Homology
Detection
Nanjiang Shu, Tuping Zhou and Sven Hovmölle
Structural Chemistry, Arrhenius Laboratory
Stockholm University
SE-106 91 Stockholm, Sweden
******@
Abstract We have developed a method for remote homology showed that sequence-sequence based methods such as BLAST
detection using profile-based fragment matching. Our method could detect most homologues with > 30% sequence identity to
compares a sliding nine-residue long fragment of the target a target sequence [3]. However, structural classifications of
sequence to all such fragments in the database and keeps up to proteins as done in SCOP (Structural Classification of Proteins)
150 candidate fragments with the highest profile-profile score. [4] and CATH (Class, Architecture, Topology and Homology
For each candidate sequence, a dot plot of the positions of the superfamily) [5] show that also proteins sharing very low
fragments in the target sequence against the positions of the sequence identities (10-20%) may still be homologous. To
matched fragments in the candidate sequence is drawn. A detect the homologues between those proteins sharing low
homology score is derived for each candidate sequence from the sequence identities, . to detect distantly related homologues,
pattern of the dot plot.
is still a challenge.
To