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

BP: Fachverband Biologische Physik

BP 18: Posters: Statistical Physics in Biological Systems

BP 18.12: Poster

Mittwoch, 28. März 2012, 17:30–19:30, Poster A

Single-molecule FRET experiments analyzed by likelihood-based methods — •Bettina G. Keller and Frank Noé — Freie Universität Berlin, Arnimallee 6, 14195 Berlin

The conformational equilibrium of single molecules can be probed by a variety of experimental techniques, such as fluorescence resonance energy transfer (FRET) spectroscopy, fluorescence quenching or atomic force experiments. However, constructing a consistent kinetic model from the measured data is difficult because (i) the conformational states of the molecule are typically related to overlapping value ranges of the observable, (ii) single-molecule experiments suffer from a low-signal to noise ratio, and (iii) the influence of measurement errors, as for example spectral cross-talk, on the signal is sizable. Likelihood-based analysis methods, such as hidden Markov models, are much more apt in dealing with overlapping states than conventional histogram analyses. However, due to problems ii and iii, their application to experimental data has remained challenging. Using a novel likelihood function, which explicitly accounts for the background noise in the experimental signal, we have constructed hidden Markov models of single-molecule FRET data of two experimental systems: the presumed two-state folder rRNA three-helix junction and the ribozyme Diels-Alderase, known to show a complex folding behavior. The analyses reveal multi-state kinetic networks including transition rates, state life times, and associated FRET efficiencies. The networks are hierarchically organized, such that at low state resolution the results of conventional histogram analyses are largely recovered.