EVOGENE Section Seminar - Bayesian integrative modeling of macromolecular complexes
Dr. Michael Habeck, Georg-August-Universität Göttingen Institute for Mathematical Stochastics, Germany
Bayesian integrative modeling of macromolecular complexes
Structural biology is shifting its focus to the study of large marcomolecular complexes. To characterize these complexes, traditional approaches such as X-ray crystallography and nuclear magnetic resonance need to be complemented by cryo-electron microscopy, cross-linking / mass spectrometry and other emerging techniques. The new data are often blurry, sparse, and noisy. It is therefore challenging to use these data for structural modeling. I will present a computational approach that addresses these challenges by means of Bayesian probability theory. I will outline the principles of Bayesian probabilistic modeling, which are applicable to biological data in general. I will illustrate our methods using various examples including a recent hybrid structure of a type 1 pilus.