Bioinformatics and Computational Molecular Biology
Algorithms and Hardware
Fuzzy Computer Code
The following Matlab M-files, input files, and results file are used to show that a PSSM-based protein domain family classifier using fuzzy-theory-based weights outperforms a linear-weighted and an unweighted classifier. The weights are formed as a combination of sequence-based conservation estimates and structure-based conservation estimates.
SurfFind.m A Matlab M-file to find residues a the surface of a protein.
Example SurfFind Input File This is a simplified version of the 1SW6 protein structure file from the Protein Data Bank (PDB).
Example SurfFind Output File This is the output from the 1SW6 file above.
Fuzzy.m A Matlab M-file to calculate the significance of family membership scores against same-family domains.
TPR input file for Fuzzy.m Includes surface/interior coding for known-structure proteins.
LRR input file for Fuzzy.m Includes surface/interior coding for known-structure proteins.
WD40 input file for Fuzzy.m Includes surface/interior coding for known-structure proteins.
Ank input file for Fuzzy.m Includes surface/interior coding for known-structure proteins.
TPR results from Fuzzy.m Using 20 randomly selected test sequences.
LRR results from Fuzzy.m Using 20 randomly selected test sequences.
WD40 results from Fuzzy.m Using 20 randomly selected test sequences.
Ank results from Fuzzy.m Using 20 randomly selected test sequences.
FuzzyOther.m A Matlab M-file to calculate the significance of family membership scores against other-family domains.
Example FuzzyOther.m Input File This is just one of the 19 other-family sequences.
All Other-Family Inputs Cut and paste a sequence into the Other.txt file.
Results from running FuzzyOther.m.
Boise State University College of Engineering
Boise State University Department of Electrical and Computer Engineering
This page created by Dr. Scott F. Smith
This page was last updated on 12 May 2004.