David Simoncini – Computational Protein Design 🗓 🗺
Guaranteed Optimization for Computational Protein Design and Applications to Bio-nanotechnology
Computational Protein Design (CPD) refers to the problem of finding a protein sequence that maximizes the stability of a protein structure, or the affinity of a protein with a binding partner. At the core of CPD, lies a compact formulation of the Global Minimum Energy Conformation problem.
This formulation assumes a rigid backbone, captures amino-acid and conformation changes through a discrete rotamer library and relies on a pairwise decomposable description of the energy. Despite its intrinsic computational hardness, CPD has started to yield several completely new functional molecules in the last decade. If deterministic algorithms give a guaranteed access to optimal designed sequences, the NP-hardness of the problem limits them to small designs. Stochastic methods are therefore often preferred. In this talk, we show how we used Artificial Intelligence optimization techniques to tackle fixed backbone, decomposable energy, rotamer based Computational Design problems with deterministic algorithms and give a few examples of applications in bio-nanotechnology.
A propos de l’intervenant: