Aller au contenu. | Aller à la navigation

Outils personnels

Navigation

Voir le monde en interaction

  • Logo CNRS
  • Logo ENSL
  • logo-enssib.jpg
Vous êtes ici : Accueil / Agenda / Séminaires / IXXI-ENS Lyon seminar

IXXI-ENS Lyon seminar

Single-Molecule Biophysics, Time Series Analysis, Information Theory.
Quand ? Le 17/06/2016,
de 11:30 à 12:30
Où ? ENS-Monod, 46 avenue d’Italie, room 116
Ajouter un événement au calendrier vCal
iCal

Hosts: Arezki Boudaoud and Claire Lesieur

Speaker: Tamiki Komatsuzaki (tamiki@es.hokudai.ac.jp), Molecule & Life Nonlinear Sciences Laboratory Research Center of Mathematics for Social Creativity Research Institute for Electronic Science,Hokkaido University, JAPAN

Abstract: Mapping experimental single-molecule time series to a Markovian network and its free energy landscape has been one of the most intriguing subjects in single molecule biophysics, especially under the constraint of finite sample data points, experimental noise inherent to the measurement, and the possible existence of memory in the process. Here we present our recently developed methods [1-5] to extract the underlying Markovian network and the effective energy landscape from time-series data. The core of the methods is the application of computational mechanics (epsilon machine) [1,2] and rate-distortion theory [3] in information theory. The epsilon machine provides a framework to derive a Markovian network from non-Markovian process. The rate-distortion theory allows the individual data points to be assigned to multiple states simultaneously, which is inevitable especially for noisy time series.

In the presentation, I briefly overview these two different methods, and demonstrate the method’s proficiency in its application to single-molecule experimental electron transfer and fluorescence resonance energy transfer trajectories obtained from biomolecules. Our method uncovers new information on hierarchical organization of states buried in the experimental trajectories that deepens our understanding of biomolecular function.

[1] Tahmina Sultana, Hiroaki Takagi, Miki Morimatsu, Hiroshi Teramoto, Chun-Biu Li, Yasushi Sako, Tamiki Komatsuzaki, “Non-Markovian properties and multiscale Hidden Markovian Network Buried in Single Molecule Time Series” J. Chem. Phys. 139, 245101 (2013)

[2] Chun-Biu Li, Haw Yang, Tamiki Komatsuzaki, “Multiscale Complex Network of Protein Conformational Fluctuation Buried in Single Molecule Time Series” Proc. Natl. Acad. Sci. USA 105, 536-541 (2008)

[3] James N. Taylor, Chun-Biu Li, David R. Cooper, Christy F. Landes, and Tamiki Komatsuzaki, “Error-based Extraction of Effective Free Energy Landscapes from Experimental Single-Molecule Time-Series” Sci. Rep. 5, 9174 (2015)

[4] Chun-Biu Li, Tamiki Komatsuzaki, “Aggregated Markov Model Using Time Series of Single Molecule Dwell Times with Minimum Excessive Information” Phys. Rev. Lett. 111,58301 (2013).

[5] Akinori Baba, Tamiki Komatsuzaki, “Construction of effective free energy landscape from single molecule time series” Proc. Natl. Acad. Sci. USA 104,19297-19302 (2007)

 

If you wish to discuss with Tamiki on the 17 of June, please contact Claire Lesieur (claire.lesieur@ens-lyon.fr)