Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
Efficient Sampling of Protein Folding Pathways using HMMSTR and Probabilistic Roadmaps Yogesh Girdhar (girdhy@cs.rpi.edu), Chris Bystroff (bystrc@rpi.edu), Srinivas Akella (sakella@cs.rpi.edu), Edward Carlson (carlse@cs.rpi.edu)
 

Summary: Efficient Sampling of Protein Folding Pathways using HMMSTR and Probabilistic Roadmaps
Yogesh Girdhar (girdhy@cs.rpi.edu), Chris Bystroff (bystrc@rpi.edu), Srinivas Akella (sakella@cs.rpi.edu), Edward Carlson (carlse@cs.rpi.edu)
Rensselaer Polytechnic Institute, Troy, New York 12180
Pathway Generation using Probabilistic Road-
maps (PRM)
Sample Generation using HMMSTR
HMMSTR [1] is a hidden Markov model for local sequence-structure correlation. It uses
knowledge of preferred orientations of amino acid sequences from data in the Protein Data
Bank (PDB) to predict the , torsion angles of local sequences. Given the amino acid se-
quence code of a protein and a window size, HMMSTR generates a set of likely , angles
for each overlapping window for the entire sequence. Over 150 motifs are modeled as
sequence-structure correlations in HMMSTR. HMMSTR uses PSI-BLAST to generate a list of
similar protein sequences. Once we have these local structures, we then proceed to build a
complete configuration out of these local structures. We start off with a random anchor
window in the protein, and then choose a random angle set (, angles for each residue
in the window) with probability proportional to its score. We then walk towards both the
ends of the protein from the anchor window, assigning angle sets for all remaining win-
dows.
Energy Minimization of HMMSTR Samples
Although HMMSTR predicts the possible local structure of an amino acid sequence, it does

  

Source: Akella, Srinivas - Department of Computer Science, University of North Carolina, Charlotte

 

Collections: Engineering; Computer Technologies and Information Sciences