An all-atom distance-dependent conditional probability discriminatory function for protein structure prediction 1 1Edited by F. Cohen

A knowledge-based scoring function based on residue triplets for protein structure prediction

“…It should be noted that a residue triplet does not necessarily consist of three residues consecutive in sequence, just as a residue pair does not necessarily correspond to a pair of neighboring residues in the two-body potential. Precisely, in terms of the Bayesian statistics formalism as described in Samudrala and Moult (1998), we view a given set of conformations for a protein sequence as comprising of two subsets: a subset of correct conformations {C} and a subset of incorrect conformations {I}. For a given conformation, we calculate the probability that it belongs to the subset of correct structures {C}, given some properties of the conformation.…”

“…As a second approach, we apply the mixture method described in Samudrala and Moult (1998), i.e. instead of using the database of the 31 500 decoy structures, the database of the 3150 solved structures is employed and averaging is done across the various residue types when determining the prior distribution.…”

“…Each of these knowledge-based functions tries to capture some aspects of the protein native conformations, such as the tendency of a certain amino acid to be exposed or buried relative to the solvent, or to be part of the helix, strand or coil local structure and so on. These knowledge-based functions are compiled based on the statistics of a database of experimentally determined protein structures (Wodak and Rooman, 1993;Sippl 1995;DeBolt and Skolnick, 1996;Gilis and Rooman, 1996;Jernigan and Bahar, 1996;Zhang et al, 1997;Samudrala and Moult, 1998). Interaction between these two categories of functions has resulted in a fertile ground for the experimentation and construction of new scoring functions.…”

Knowledge-Based Energy Functions for Computational Studies of Proteins

“…In the work of developing the "Residue-specific All-atom Probability Discriminatory Function" (RAPDF) (Samudrala and Moult, 1998), Samudrala and Moult (1998) alternatively replaced Equation 4.9 with:…”

“…Because it is limited by the simple spatial description of a cut-off distance, it cannot capture the finer spatial details. Several distancedependent potentials have been developed to overcome this limitation, and in general have better performance (Lu and Skolnick, 2001;Samudrala and Moult, 1998;Zhou and Zhou, 2002). A major focus of works in this area is the development of models for the reference state.…”