DISCOTRESS - Revision history

Djs244 at 10:37, 29 September 2021

2021-09-29T10:37:41Z

← Older revision		Revision as of 10:37, 29 September 2021
Line 8:		Line 8:

	* calculate the probability of visiting states on a direct transition path between two endpoint states		* calculate the probability of visiting states on a direct transition path between two endpoint states

			* calculate the committor probabilities for all nodes of the network (and hence analyze the network by equilibrium or nonequilibrium transition path theory)

	* decompose the ensemble of first passage paths into a finite set of flux-paths and their associated contributions to the net productive flux		* decompose the ensemble of first passage paths into a finite set of flux-paths and their associated contributions to the net productive flux
Line 24:		Line 26:

	Further helpful scripts to analyze the output of '''DISCOTRESS''' are available [https://github.com/danieljsharpe/DISCOTRESS_tools here].		Further helpful scripts to analyze the output of '''DISCOTRESS''' are available [https://github.com/danieljsharpe/DISCOTRESS_tools here].

			==Analyzing your database==

			The routines that are implemented in DISCOTRESS allow for numerous detailed analyses of kinetic networks that are not possible by conventional methods. You may wish to conduct a particular analysis to complement your initial report of the database and validate your qualitative observations based on the topology of the energy landscape. Alternatively, you may wish to conduct an extensive quantitative analysis of a database that you have constructed in a separate report. Using DISCOTRESS, you can e.g. find the key transition paths that facilitate the productive transition between endpoint states (by shortest paths analysis, transition path theory, or simulation of trajectories) and identify the key states that mediate these pathways (by calculating reactive visitation probabilities, committor probabilities, etc.)

			A presentation ("Sharpe - Recent advances in the numerical analysis of finite Markov chains") describing the quantitative methods is available in the "AVT presentations" folder of the Wales group google drive [https://drive.google.com/drive/u/0/folders/0ByQUHkoP5DEnOVhUR3RzTUQyWmc?resourcekey=0-zfSotsYqPBhxA3KZPiuuDg here].



	==Tutorials==		==Tutorials==

Djs244 at 11:30, 1 July 2021

2021-07-01T11:30:33Z

← Older revision		Revision as of 11:30, 1 July 2021
Line 1:		Line 1:
	'''DISCOTRESS''' is a C++ program to simulate and analyze the dynamics on kinetic transition networks, beyond the functionality provided in [[PATHSAMPLE]]. '''DISCOTRESS''' is written and maintained by Daniel J. Sharpe, and can be found here		'''DISCOTRESS''' is a C++ program to simulate and analyze the dynamics on kinetic transition networks, beyond the functionality provided in [[PATHSAMPLE]]. '''DISCOTRESS''' is written and maintained by Daniel J. Sharpe, and can be found [https://github.com/danieljsharpe/DISCOTRESS here].

	==Functionality==		==Functionality==
Line 7:		Line 7:
	* simulate trajectories by kinetic path sampling, thereby sampling the first passage time distribution and obtaining a numerical estimate for the time-dependent occupation probability distribution of macrostates		* simulate trajectories by kinetic path sampling, thereby sampling the first passage time distribution and obtaining a numerical estimate for the time-dependent occupation probability distribution of macrostates

	* calculate the probability of visiting ~~a state~~ on a direct transition path between two endpoint states		* calculate the probability of visiting states on a direct transition path between two endpoint states

	* decompose the ensemble of first passage paths into a finite set of flux-paths and their associated contributions to the net productive flux		* decompose the ensemble of first passage paths into a finite set of flux-paths and their associated contributions to the net productive flux
Line 13:		Line 13:
	==Getting Started==		==Getting Started==

	'''DISCOTRESS''' has its own system of input files, so first you will need to convert your kinetic network database from '''PATHSAMPLE''' format (''min.data'',''ts.data''). You will need to construct the following 3 files:		'''DISCOTRESS''' has its own system of input files, so first you will need to convert your kinetic network database from '''PATHSAMPLE''' format (''min.data'', ''ts.data''). You will need to construct the following 3 files:

	* ''edge_conns.dat'' - two-column file containing a list of connections, can be extracted straightforwardly as the corresponding columns of ''ts.data''		* ''edge_conns.dat'' - two-column file containing a list of connections, can be extracted straightforwardly as the corresponding columns of ''ts.data''
Line 21:		Line 21:
	* ''stat_prob.dat'' - single-column file containing the natural log of stationary probabilities of nodes in the network. You will need to dump these data, which are directly related to the energies of minima, from '''PATHSAMPLE''' (Fortran script available).		* ''stat_prob.dat'' - single-column file containing the natural log of stationary probabilities of nodes in the network. You will need to dump these data, which are directly related to the energies of minima, from '''PATHSAMPLE''' (Fortran script available).

	Further information on these '''DISCOTRESS''' input files and others, including the ''input.kmc'' file of keywords, is available here		Further information on these '''DISCOTRESS''' input files and others, including the ''input.kmc'' file of keywords, is available [https://github.com/danieljsharpe/DISCOTRESS/blob/master/documentation.md here].

			Further helpful scripts to analyze the output of '''DISCOTRESS''' are available [https://github.com/danieljsharpe/DISCOTRESS_tools here].

	==Tutorials==		==Tutorials==

	Tutorials are available here		Tutorials are available [https://github.com/danieljsharpe/DISCOTRESS_tutorials here].

			==FAQs==

			Answers to some FAQs are available [https://github.com/danieljsharpe/DISCOTRESS/blob/master/FAQs.md here].

Djs244: Created DISCOTRESS page

2021-07-01T11:20:10Z

Created DISCOTRESS page

New page

'''DISCOTRESS''' is a C++ program to simulate and analyze the dynamics on kinetic transition networks, beyond the functionality provided in [[PATHSAMPLE]]. '''DISCOTRESS''' is written and maintained by Daniel J. Sharpe, and can be found here

==Functionality==

'''DISCOTRESS''' contains implementations of state-of-the-art algorithms to robustly and efficiently perform quantitative analyses of a kinetic network. Some recommended functionality that is very powerful for gaining a detailed understanding of your kinetic network is as follows:

* simulate trajectories by kinetic path sampling, thereby sampling the first passage time distribution and obtaining a numerical estimate for the time-dependent occupation probability distribution of macrostates

* calculate the probability of visiting a state on a direct transition path between two endpoint states

* decompose the ensemble of first passage paths into a finite set of flux-paths and their associated contributions to the net productive flux

==Getting Started==

'''DISCOTRESS''' has its own system of input files, so first you will need to convert your kinetic network database from '''PATHSAMPLE''' format (''min.data'',''ts.data''). You will need to construct the following 3 files:

* ''edge_conns.dat'' - two-column file containing a list of connections, can be extracted straightforwardly as the corresponding columns of ''ts.data''

* ''edge_weights.dat'' - two-column file containing the natural log of transition rates corresponding to the entries in ''edge_conns.dat''. You will need to dump these data from '''PATHSAMPLE''' (Fortran script available)

* ''stat_prob.dat'' - single-column file containing the natural log of stationary probabilities of nodes in the network. You will need to dump these data, which are directly related to the energies of minima, from '''PATHSAMPLE''' (Fortran script available).

Further information on these '''DISCOTRESS''' input files and others, including the ''input.kmc'' file of keywords, is available here

==Tutorials==

Tutorials are available here