WELCOME TO TMhhcp PAGE!

TMhhcp is developed to predict residue-residue contacts in alpha-helix transmembrane proteins. The predicted contacts could be further used to predict helix-helix interactions between helices.

Up to now, it's still difficult and expensive to experimentally determine transmembrane protein structures. Therefore, an accurate computational method for predicting the structure of the transmembrane protein is desirable. Predicted residue-residue contacts could be an important constraint to the membrane protein structure prediction.

In this server, we provide two prediction models corresponding to two residue contact definitions. The user could choose any one to predict contact residue pairs for the submitted protein.

TMhhcp uses the machine learning method, random forest, which integrates PSSM, sequence distance, residue covariance, residue conservation, relative residue distance of two helices and helix number to predict helix-helix contacts. Through our test, TMhhcp outperforms two previous helix-helix contacts predictors on accuracy.

TMhhcp is easy to use and the required inputs are the query transmembrane protein sequence and the corresponding topology information. It's hoped that TMhhcp can be helpful to the community. In case any question, please feel free tocontact Xiao-Feng Wang.