Almost FPRAS for Lattice Models of Protein Folding
The provably efficient randomized approximation scheme which evaluates the partition function for the wide class of lattice models of protein prediction is presented. We propose to apply the idea of self-testing algorithms introduced recently in . We consider the protein folding process which is simplified to a self-avoiding walk on a lattice. The power of a simplified approach is in its ability to search the conformation space, to train the search parameters and to test basic assumptions about the nature of the protein folding process. Our main theoretical results are formulated in the general setting, i.e. we do not assume any specific lattice model. For the simulation study we have chosen the HP model on the FCC lattice.
KeywordsMarkov Chain Partition Function Lattice Model Protein Structure Prediction Markov Chain Monte Carlo Algorithm
Unable to display preview. Download preview PDF.
- 3.Madras, N., Slade, G.: The Self-Avoiding Walk, Birkhäuser, Boston (1993)Google Scholar
- 5.Niemiro, W., Pokarowski, P.: Faster MCMC Estimation Along One Walk (2001) (Preprint)Google Scholar
- 6.Pokarowski, P., Koliński, A., Skolnik, J.: A minimal physically realistic protein-like lattice model: Designing an enargy landscape that ensures all-or-one folding to a unique native state. Biophysical Journal (2003) (in press)Google Scholar
- 7.Randall, D.: Counting in lattices: combinatorial problems from statistical mechanics. PhD Thesis, UC Berkeley (1995)Google Scholar
- 9.Schuster, P., Stadler, P.F.: Discrete Models of Biopolymers (1999) (TBI Preprint)Google Scholar
- 10.Sinclair, A.J.: Algorithms for random generation and counting: a Markov chain approach, Birkhäuser, Boston (1992)Google Scholar
- 11.Will, S.: Constraint-based hydrophobic core construction for protein structure prediction in the face-centered-cubic lattice. In: Proc. Pacific Symposium on Biocomputing (2002)Google Scholar