Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

BP: Fachverband Biologische Physik

BP 35: SYBE: Statistical Physics and Biological Evolution

BP 35.2: Hauptvortrag

Freitag, 18. März 2011, 11:00–11:30, TRE Ma

Correlated mutations: Facts or artifacts? — •Amnon Horovitz — Weizmann Institute, Rehovot, Israel

Mutations that affect protein function by structural perturbation at one site are often compensated for by mutations at other sites. Such correlated mutations are thought to occur since there is greater selective pressure to conserve protein structure and function than sequence. Correlated mutation analyses have indicated that distant sites in proteins are often coupled to each other. It has not been clear, however, whether such correlations between distant positions reflect real long-range interactions or common ancestry. In order to address this question, lattice models of proteins were subjected to mutation and selection for greater stability and long-range correlations that arose as a result were characterized. Our results show that long-range correlations with non-zero coupling energies do exist in lattice models [1] and that they are more common when the stability of the native state is achieved by negative design, i.e. by destabilizing non-native contacts [2]. The implications of these findings for real proteins will be discussed.

[1] O. Noivirt-Brik, R. Unger and A. Horovitz, BMC Struct. Biol. 9, 4 (2009).

[2] O. Noivirt-Brik, A. Horovitz, and R. Unger, PLoS Comput. Biol. 5, e1000592 (2009).