Markov State Models Gregory R. Bowman Miller Fellow University of California, Berkeley Tuesday, August 16, 2011
Defining The Protein Folding Problem
Defining The Protein Folding Problem
Why Mechanism?
Why Mechanism?
Why Mechanism?
The Sampling Challenge Bond vibration Isomerization Water dynamics Helix forms Fast folding Slow conf change protein oligomerization 10-15 femto 10-12 pico 10-9 nano 10-6 micro 10-3 milli 10 0 seconds 10 3 seconds MD step long MD run where we d love to be time on 1 fast CPU 1 day 3 years 3,000 years 3,000,000 years ~age of the earth Tuesday, August 16, 2011
The Sampling Challenge Bond vibration Isomerization Water dynamics Helix forms Fast folding Slow conf change protein oligomerization 10-15 femto 10-12 pico 10-9 nano 10-6 micro 10-3 milli 10 0 seconds 10 3 seconds MD step long MD run where we d love to be time on 1 fast CPU 1 day 3 years 3,000 years 3,000,000 years ~age of the earth Tuesday, August 16, 2011
The Sampling Challenge Bond vibration Isomerization Water dynamics Helix forms Fast folding Slow conf change protein oligomerization 10-15 femto 10-12 pico 10-9 nano 10-6 micro 10-3 milli 10 0 seconds 10 3 seconds MD step long MD run where we d love to be time on 1 fast CPU 1 day 3 years 3,000 years 3,000,000 years ~age of the earth Tuesday, August 16, 2011
The Sampling Challenge Bond vibration Isomerization Water dynamics Helix forms Fast folding Slow conf change protein oligomerization 10-15 femto 10-12 pico 10-9 nano 10-6 micro 10-3 milli 10 0 seconds 10 3 seconds MD step long MD run where we d love to be time on 1 fast CPU 1 day 3 years 3,000 years 3,000,000 years ~age of the earth Tuesday, August 16, 2011
Markov State Models (MSMs)
Markov State Models (MSMs)
Markov State Models (MSMs)
How to Build an MSM
How to Build an MSM
How to Build an MSM
Sample
Sample
Sample
Sample
Clustering Gives a High- resolution Model (this is a cartoon, since it s hard to draw 10,000 states in a talk) Tuesday, August 16, 2011
Lumping Provides Human Intuition
Under the Hood!"# $!"!% $!"& $'$ T(! t) =!"& $!"( $!"& $'$! $ $!"!) $!"* $'$ ' $ $' $ $' $ $'$!"#$%&#'&(&)"%&#'!"#'& Tuesday, August 16, 2011
Free Energy (kt) Modeling Experiments native state prediction Bowman, Beauchamp, Boxer, and Pande. JCP 2009. Bowman, Voelz, and Pande. JACS 2011. Tuesday, August 16, 2011
Free Energy (kt) Modeling Experiments native state prediction Bowman, Beauchamp, Boxer, and Pande. JCP 2009. Bowman, Voelz, and Pande. JACS 2011. Tuesday, August 16, 2011
Free Energy (kt) Modeling Experiments native state prediction!"#$%&#'&(&)"%&#'!"#'& Bowman, Beauchamp, Boxer, and Pande. JCP 2009. Bowman, Voelz, and Pande. JACS 2011. *+,"#'&(&!"#'!*+,& Tuesday, August 16, 2011
One Area Where We Need Help...
One Area Where We Need Help...
One Area Where We Need Help...
Jump- starting an MSM with Rosetta
Jump- starting an MSM with Rosetta
Jump- starting an MSM with Rosetta
What Can We Do For You?
What Can We Do For You? Dynamics in Design Tuesday, August 16, 2011
What Can We Do For You? Dynamics in Design Sampling Tuesday, August 16, 2011
What Can We Do For You? Dynamics in Design Sampling Force Field Assessment Tuesday, August 16, 2011
What Can We Do For You? Dynamics in Design Sampling Force Field Assessment? Tuesday, August 16, 2011
Conclusions
Conclusions Mechanism is important Tuesday, August 16, 2011
Conclusions Mechanism is important Markov state models are a powerful way of capturing the mechanisms of conformational change Tuesday, August 16, 2011
Conclusions Mechanism is important Markov state models are a powerful way of capturing the mechanisms of conformational change There s lots of room for collaboration! Tuesday, August 16, 2011
Thanks!
Conclusions Mechanism is important Markov state models are a powerful way of capturing the mechanisms of conformational change There s lots of room for collaboration! Tuesday, August 16, 2011