ITR: A Protocol For Computational Protein Design

点击翻译按钮获取中文摘要

This award was made on a 'small' category proposal submitted in response to the ITR solicitation, NSF-02-168. It supports computational research on protein folding and design. The PIs aim to provide a comprehensive computational protocol utilizing the principle of designability in order to address the fundamental problem of designing novel protein folds. Algorithmic improvements will be made to enhance the modulus on backbone selection in the protocol. Research will address issues ranging from backbone modeling to structure screening and clustering to stochastic computation of designability to obtain an optimal strategy in the search for novel backbone conformations of much longer proteins than previously thought possible. A high-resolution all-atomic modeling will be employed to pack suitable choices of side-chains onto the backbone to obtain realistic protein conformations for experimental verification. New elements that will be actively explored, include: (1) a more accurate pairwise decomposition of the solvation energy function; (2) improving a widely used method of side-chain packing to obtain a significant increase of computational speed for the deterministic dead-end-elimination algorithm with stochastic sampling schemes embedded; and (3) a novel parametric scheme to characterize the flexibilities of secondary structure elements of a protein structure to overcome the severe limitation of the fixed-backbone methodology imposed by most of currently existing algorithms. This award also supports education at the graduate and postdoctoral level and provides research experience in an industrial laboratory. %%%This award was made on a 'small' category proposal submitted in response to the ITR solicitation, NSF-02-168. It supports multidisciplinary computational research involving statistical mechanics, molecular biophysics and computer science on protein folding and design. While there is a great diversity of amino acid sequences in the many proteins observed in nature, about 100,000 different proteins in the human body alone, the diversity of structures is much more limited. It is estimated that there are only about 1000 distinct protein folds in nature. The PIs seek to develop a method for designing novel protein folds that is based on the "designability principle." The designability principle, developed by the PIs, connects highly designable structures and highly stable amino-acid sequences. Research involves implementing and improving computational algorithms. By a focused and methodical examination of various options, criteria, and algorithms at the different stages of the proposed protocol, the PIs aim to design a powerful and practical tool for protein design by computer. This award also supports education at the graduate and postdoctoral level and provides research experience in an industrial laboratory. ***