Protein Folding And Self Avoiding Walks Polyhedral Studies And Solutions

Protein Folding and Self-Avoiding Walks Polyhedral Studies and Solutions

The protein folding problem refers to the correlation of a protein's amino acid sequence and its native three-dimensional structure which is essential for functionality. It still constitutes one of the major challenges in computational biology. One commonly studied model for the protein folding problem is the HP lattice model in which proteins are considered in a fairly abstract representation. However, the HP model proteins exhibit significant parallels to proteins occurring in nature. The solution of the HP lattice mode as a combinatorial optimization problem has been proven to be NP-complete, and there have already been developed various different approaches for efficient algorithms. We study an integer programming formulation of the problem. Starting with an analysis of this model, where we concentrate on symmetry issues, we show how the model can be consolidated by exploiting symmetry properties of the underlying lattice. The main focus lies in the development of specific components of a branch-and-cut framework for the computation of solutions for the HP model by means of integer programming methods. In order to understand the structure of the model, we perform a series of polyhedral studies from which we derive two main classes of cutting planes. Furthermore, we exploit the knowledge of folding principles which are also valid for HP model proteins for the development of related branching strategies. For the solution of a special class of instances, we present an implementation of a genetic algorithm for the generation of primal feasible start solutions. Finally, we document the performance of the methods developed for each of the four topics (model consolidation, primal method, branching strategy and cutting planes) within the branch-and-cut procedure. We present computational results for different types of lattices, where we both consider known benchmark instances from literature and random instances.

Facets of Combinatorial Optimization

Martin Grötschel is one of the most influential mathematicians of our time. He has received numerous honors and holds a number of key positions in the international mathematical community. He celebrated his 65th birthday on September 10, 2013. Martin Grötschel’s doctoral descendant tree 1983–2012, i.e., the first 30 years, features 39 children, 74 grandchildren, 24 great-grandchildren and 2 great-great-grandchildren, a total of 139 doctoral descendants. This book starts with a personal tribute to Martin Grötschel by the editors (Part I), a contribution by his very special “predecessor” Manfred Padberg on “Facets and Rank of Integer Polyhedra” (Part II), and the doctoral descendant tree 1983–2012 (Part III). The core of this book (Part IV) contains 16 contributions, each of which is coauthored by at least one doctoral descendant. The sequence of the articles starts with contributions to the theory of mathematical optimization, including polyhedral combinatorics, extended formulations, mixed-integer convex optimization, super classes of perfect graphs, efficient algorithms for subtree-telecenters, junctions in acyclic graphs and preemptive restricted strip covering, as well as efficient approximation of non-preemptive restricted strip covering. Combinations of new theoretical insights with algorithms and experiments deal with network design problems, combinatorial optimization problems with submodular objective functions and more general mixed-integer nonlinear optimization problems. Applications include VLSI layout design, systems biology, wireless network design, mean-risk optimization and gas network optimization. Computational studies include a semidefinite branch and cut approach for the max k-cut problem, mixed-integer nonlinear optimal control, and mixed-integer linear optimization for scheduling and routing of fly-in safari planes. The two closing articles are devoted to computational advances in general mixed integer linear optimization, the first by scientists working in industry, the second by scientists working in academia. These articles reflect the “scientific facets” of Martin Grötschel who has set standards in theory, computation and applications.

Deutsche Nationalbibliographie und Bibliographie der im Ausland erschienenen deutschsprachigen Veröffentlichungen