Center for Bioinformatics at KU

As part of our public Seminar Series
http://bioinformatics.ku.edu/seminars

September 9, Tue 2008
11:00 am, MRB 200 Conference Room

Dr. Sitao Wu

Zhang Lab, University of Kansas Center for Bioinformatics,

Sequence-based protein contact prediction and its usage in 3D protein structure modeling

To reduce the conformation space, currently template-based contact predictions are widely used in 3D protein structure predictions. These contacts are used as spatial restraints to guide 3D structure simulations. For “hard” targets without identified good templates, the accuracy of template-based contact predictions is relatively too low to generate a model with good topology. On the other hand, we found that sequence-based contact predictions can provide enough new contact information complementary to template-based predictions even if the accuracy of them is still low for “hard” targets. By injecting the sequence-based contact prediction into Monte Carlo simulation of protein structures (I-TASSER), we find that there is some improvement of final models of hard targets.