franknoe
commented
6 years ago Am 25/06/18 um 11:46 schrieb Eric Lang:
Hello,
/This is a general technical question, not a specific issue with PyEMMA./
I am simulating a system transitioning between an open and closed state. I have initiated several simulations from the open and from the closed crystal structures and observed a number of transitions over the course of the simulations. I would like to create a MSM to study this transition between states.
Since I have simulations starting from the open state and simulations starting from the closed state, when concatenated together into a single trajectory, the end of an open form simulation joint to a the beginning of a closed form simulation, for example, will look like a transition where in fact there is none.
Is there a way to avoid this in PyEMMA?
Of course. Just don't concatenate simulations.
For example, lets suppose I have two trajectories, one starting from the closed form and one starting from the open form and in each of those trajectory I do not observe any transition to the other state. If I do something like:
|>>> from pyemma.coordinates import load >>> files = ['start_from_open_1.xtc', 'start_from_closed_1.xtc'] >>> output = load(files, top='my_structure.pdb') |
and try to build a MSM from there, will PyEMMA believe that there is one transition between states (because the simulations sample two different states), or will it correctly found that there is no transition between the two states?
It will correctly treat them as separate trajectories. That is the whole point of programs like PyEMMA because it is the typical case in MSM-building that you have many independently run short trajectories.
BTW, if you do a "load" you will literally load everything in memory. This only works if you don't have too much data of course. PyEMMA also allows you to stream data on the fly, so that you never have to load everything. Maybe this is not an issue, but if it is, replace load by source, and just inject the resulting object into the subsequent analysis tools, such as TICA.
Best, Frank.
Many thanks in advance for your help,
Eric
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/markovmodel/PyEMMA/issues/1321, or mute the thread https://github.com/notifications/unsubscribe-auth/AGMeQuFQwjshY36sGAMHACTtHThj3CuUks5uALFagaJpZM4U11Ph.
--
Prof. Dr. Frank Noe Head of Computational Molecular Biology group Freie Universitaet Berlin
eric-jm-lang
Hello,
This is a general technical question, not a specific issue with PyEMMA.
I am simulating a system transitioning between an open and closed state. I have initiated several simulations from the open and from the closed crystal structures and observed a number of transitions over the course of the simulations. I would like to create a MSM to study this transition between states.
Since I have simulations starting from the open state and simulations starting from the closed state, when concatenated together into a single trajectory, the end of an open form simulation joint to a the beginning of a closed form simulation, for example, will look like a transition where in fact there is none.
Is there a way to avoid this in PyEMMA?
For example, lets suppose I have two trajectories, one starting from the closed form and one starting from the open form and in each of those trajectory I do not observe any transition to the other state. If I do something like:
and try to build a MSM from there, will PyEMMA believe that there is one transition between states (because the simulations sample two different states), or will it correctly found that there is no transition between the two states?
Many thanks in advance for your help,
Eric