Protein structure prediction based on fragment assembly and the beta-strand pairing energy function

초록

We develop an improved version of PROFESY, a novel method for ab-initio prediction of protein tertiary structures based on fragment assembly and global optimization. In contrast to the primitive version presented earlier, where the hydrogen bond was defined only in terms of inter-atom distance, the angle dependence is now correctly incorporated. This new feature allows us to obtain low-energy conformations with a reasonable number of beta strands, in contrast to the earlier version in which the fraction of alpha helices was excessively large on average. In order to enhance the performance of the prediction method, we optimize the linear parameters of the energy function so that native-like conformations become energetically more favorable than non-native ones for proteins with known structures. We test the feasibility of the parameter optimization procedure by applying it to a training set consisting of two proteins of structural class alpha + beta: 1FSD and 1PQS. We. use the resulting parameter set for the jackknife test on several proteins from various structural classes. The results are quite promising. In particular, for protein 2GB1, the predictions improve dramatically with the optimized parameter set compared to the original parameters, despite the fact that 2GB1 was not included in the training set. This result suggests that parameters trained for a relatively small number of proteins are transferrable to other proteins to some extent.

키워드