We present an exact method for the global minimum energy conformation (GMEC) search of protein side-chains. Our method consists of a branch-and-bound (B&B) framework and a new subproblempruning scheme. The pruning scheme consists of upper/lower-bounding methods and problem-size reduction techniques. We explore a way of using the tree-reweighted max-product algorithm for computing lowerbounds of the GMEC energy. The problem-size reduction techniques are necessary when the size of the subproblem is too large to rely on more accurate yet expensive bounding methods. The experimental results show our pruning scheme is effective and our B&B method exactly solves protein sequence design cases that are very hard to solve with the dead-end elimination. © Springer-Verlag Berlin Heidelberg 2006.
CITATION STYLE
Hong, E. J., & Lozano-Pérez, T. (2006). Protein side-chain placement through MAP estimation and problem-size reduction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4175 LNBI, pp. 219–230). Springer Verlag. https://doi.org/10.1007/11851561_21
Mendeley helps you to discover research relevant for your work.