Software

The Computational Molecular Biology Lab develops software for the simulation of molecular processes.

Our principles for scientific software development are

  • clean design and reusable code
  • ensuring correct functionality by unit and integration tests
  • extensive documentation and sample notebooks
  • automatic testing and deployment of each release on all target platforms
  • open-source policy and development via Github
  • workshops to provide hands-on training to users  

We believe that the benefits of this effort outweigh the time and costs involved to maintain it: Software development and dissemination will increase the impact of developed methods,  help to maintain group knowledge, serve the community and ensure reproducibility of results. 

Collapse allExpand all

PyEMMA is a Python library for the estimation, validation, and analysis of kinetic models from MD data. Functionalities include dimension reduction techniques such as the time-lagged independent component analysis, clustering, maximum-likelihood and Bayesian estimation of Markov State Models and Hidden Markov Models, coarse-graining and analysis of kinetic models, computation of transition pathways, and visualization tools.

Code: github.com/markovmodel/pyemma
Languages: Python (90%), C (10%)
Usage: > 50,000 downloads
License: Lesser GPL (open-source)
Release: 2.4, 34 releases
Platforms: Linux, Windows, Mac
DOI: 10.1021/acs.jctc.5b00743
DOI: 10.1021/ct300274u



Direct link

ReaDDy: High-performance simulator for reaction-diffusion dynamics with interacting particles. Suitable for modeling cellular signal transduction, chemical microreactors, soft matter systems, membrane mechanics, virus assembly etc. Previous version was Java-based. New version currently developed with software abstraction layers: (i) Python user API, (ii) C++ API, (iii) Multithreading/CUDA/MPI kernels for execution of different platforms.

Code: github.com/readdy/readdy
Languages: C++ (70%) CUDA (20%) Python (10%)
Usage: ca. 1,000 downloads
License: Lesser GPL (open-source)
Release: 1.0 (Java), in preparation (C++)
Platforms: Linux, Mac
DOI: 10.1371/journal.pone.0074261
DOI: 10.1016/j.bpj.2014.12.025

Direct link

Languages C++ (70%) CUDA (20%) Python (10%) License Lesser GPL (open-source)

Release 1.0 (Java), in preparation (C++) Platforms Linux, Mac

msmtools

Library Markov chain inference Python/C, LGPL license. v1.2.1 (12 releases).
https://github.com/markovmodel/msmtools

bhmm

Library for Hidden Markov model inference Python/C, LGPL license. v0.5.2 (9 releases).
https://github.com/bhmm/bhmm

thermotools 

Library for statistical inference from multi-ensemble data, Python/C, LGPL license. v0.2.6 (22 releases).
https://github.com/markovmodel/thermotools

deeptime

Deep learning for time series data Python, in development. 

Direct link