TransMEP - Transfer learning for Mutant Effect Prediction

TransMep is a tool for using transfer learning embeddings from protein language models to train variant prediction models from existing mutagenesis data. It is focused on speed and simplicity of use. You just input your dataset and obtain a prediction model, accompanied by detailed reports on performance, hyperparameter optimization, training samples importance and even an attribution to individual mutations. Get the software from GitHub and learn more about it in our publication.

ATRANET: Automated TRAnsition NETwork generation

ATRANET is a Python program that allows the automated generation of transition networks from molecular dynamics data of proteins. It can treat individual proteins and aggregating peptides. Moreover, the selection of the descriptors for the definition of the transition states – the nodes of the tranisitin network – is guided by a correlation analysis between different descriptors. The program is available at and more information can be found in Methods 206, 18-26, (2022).

TICAgg: TICA for aggregating systems

TICAgg is a Python notebook that resolves the degeneracy problem of molecular assemblies by sorting the permutable molecular distances. This sorting approach was combined with the dimensionality reduction technique TICA (time-lagged independent component analysis) and tested for Markov state models of peptide aggregation. The notebook is available at and more information can be found in J. Chem. Phys., 150: 115101 (2019) and bioRxiv, DOI:10.1101/2020.04.25.060269 (2020).

MOPS2: Molecular Order Parameters S2

MOPS2 is Python program that allows to compute bond vector S2 order parameters from MD trajectories. The software and detailed installation instructions are available at
See J. Phys. Chem. B, 123: 1453-1480 (2019) for more information.

FACTSMEM: a Generalized Born implicit solvent model for membranes

Information on the theory and implementation can be found here:
J. Chem. Theory Comput., 10, 3163-3176 (2014)

Zn2+ and Cu2+ dummy models for classical MD simulations

The topology and parameter files for nonbonded dummy models for Zn2+ and Cu2+ as published in Development and Application of a Nonbonded Cu2+ Model That Includes the Jahn-Teller Effect by Q. Liao, S.C.L. Kamerlin, and B. Strodel, J. Phys. Chem. Lett., 6, 2657-2662 (2015) are provided:

Cationic dummy models with ion-induced dipole interactions for MD simulations

The topology and parameter files for nonbonded dummy models for Mg2+, Al3+, Fe3+, and Cr3+ as published in Extending the Nonbonded Cationic Dummy Model to Account for Ion-Induced Dipole Interactions by Q. Liao, A. Pabis, B. Strodel, and S.C.L. Kamerlin, J. Phys. Chem. Lett., 8, 5408-5414 (2017) are provided: