CWL AutoEncoders for MD Analysis tutorial
Version 1

View on GitHub Download RO-Crate
Workflow Type: Common Workflow Language
Stable

AutoEncoders for Anomaly Detection tutorial using BioExcel Building Blocks (biobb)

This tutorial involves the use of a multilayer AutoEncoder (AE) for feature extraction and pattern recognition by analyzing Molecular Dynamic Simulations, step by step, using the BioExcel Building Blocks library (biobb).

Copyright & Licensing

This software has been developed in the MMB group at the BSC & IRB for the European BioExcel, funded by the European Commission (EU Horizon Europe 101093290, EU H2020 823830, EU H2020 675728).

Licensed under the Apache License 2.0, see the file LICENSE for details.

Click and drag the diagram to pan, double click or use the controls to zoom.

SEEK ID: https://workflowhub.eu/workflows/2144?version=1

DOI: 10.48546/workflowhub.workflow.2144.1

Inputs

ID Name Description Type
step1_gmx_image1_input_traj_path Input file Path to the GROMACS trajectory file.
  • File
step1_gmx_image1_input_top_path Input file Path to the GROMACS input topology file.
  • File
step1_gmx_image1_output_traj_path Output file Path to the output file.
  • string
step1_gmx_image1_config Config file Configuration file for biobb_analysis.gmx_image tool.
  • string
step2_mdfeaturizer1_input_topology_path Input file Path to the input topology file.
  • File
step2_mdfeaturizer1_output_dataset_pt_path Output file Path to the output dataset model file.
  • string
step2_mdfeaturizer1_output_stats_pt_path Output file Path to the output model statistics file.
  • string
step2_mdfeaturizer1_config Config file Configuration file for biobb_pytorch.mdfeaturizer tool.
  • string
step3_build_model_output_model_pth_path Output file Path to save the model in .pth format.
  • string
step3_build_model_config Config file Configuration file for biobb_pytorch.build_model tool.
  • string
step4_train_model_output_model_pth_path Output file Path to save the trained model (.pth). If omitted, the trained model is only available in memory.
  • string
step4_train_model_output_metrics_npz_path Output file Path save training metrics in compressed NumPy format (.npz).
  • string
step4_train_model_config Config file Configuration file for biobb_pytorch.train_model tool.
  • string
step5_gmx_image2_input_traj_path Input file Path to the GROMACS trajectory file.
  • File
step5_gmx_image2_input_top_path Input file Path to the GROMACS input topology file.
  • File
step5_gmx_image2_output_traj_path Output file Path to the output file.
  • string
step5_gmx_image2_config Config file Configuration file for biobb_analysis.gmx_image tool.
  • string
step6_mdfeaturizer2_input_topology_path Input file Path to the input topology file.
  • File
step6_mdfeaturizer2_output_dataset_pt_path Output file Path to the output dataset model file.
  • string
step6_mdfeaturizer2_output_stats_pt_path Output file Path to the output model statistics file.
  • string
step6_mdfeaturizer2_config Config file Configuration file for biobb_pytorch.mdfeaturizer tool.
  • string
step7_evaluate_model_output_results_npz_path Output file Path to the output evaluation results file (compressed NumPy archive).
  • string
step7_evaluate_model_config Config file Configuration file for biobb_pytorch.evaluate_model tool.
  • string
step8_make_ndx1_input_structure_path Input file Path to the input GRO/PDB/TPR file.
  • File
step8_make_ndx1_output_ndx_path Output file Path to the output index NDX file.
  • string
step8_make_ndx1_config Config file Configuration file for biobb_gromacs.make_ndx tool.
  • string
step9_make_ndx2_input_structure_path Input file Path to the input GRO/PDB/TPR file.
  • File
step9_make_ndx2_output_ndx_path Output file Path to the output index NDX file.
  • string
step9_make_ndx2_config Config file Configuration file for biobb_gromacs.make_ndx tool.
  • string
step10_gmx_rmsf1_input_structure_path Input file Path to the input structure file.
  • File
step10_gmx_rmsf1_input_traj_path Input file Path to the GROMACS trajectory file.
  • File
step10_gmx_rmsf1_output_xvg_path Output file Path to the XVG output file.
  • string
step10_gmx_rmsf1_config Config file Configuration file for biobb_analysis.gmx_rmsf tool.
  • string
step11_gmx_rmsf2_input_structure_path Input file Path to the input structure file.
  • File
step11_gmx_rmsf2_input_traj_path Input file Path to the GROMACS trajectory file.
  • File
step11_gmx_rmsf2_output_xvg_path Output file Path to the XVG output file.
  • string
step11_gmx_rmsf2_config Config file Configuration file for biobb_analysis.gmx_rmsf tool.
  • string
step12_feat2traj_output_traj_path Output file Path to save the trajectory in xtc/pdb/dcd format.
  • string
step12_feat2traj_output_top_path Output file Path to save the output topology file (pdb). Used if trajectory format requires separate topology.
  • string
step13_gmx_rmsf3_output_xvg_path Output file Path to the XVG output file.
  • string
step13_gmx_rmsf3_config Config file Configuration file for biobb_analysis.gmx_rmsf tool.
  • string
step14_make_plumed_input_reference_pdb_path Input file Path to reference PDB used for FIT_TO_TEMPLATE actions when Cartesian features are present.
  • File
step14_make_plumed_output_model_ptc_path Output file Path to the output TorchScript model file (.ptc) for PLUMED's PYTORCH_MODEL action.
  • string
step14_make_plumed_output_plumed_dat_path Output file Path to the output PLUMED input file.
  • string
step14_make_plumed_output_features_dat_path Output file Path to the output features.dat file describing the CVs to PLUMED.
  • string
step14_make_plumed_config Config file Configuration file for biobb_pytorch.make_plumed tool.
  • string

Steps

ID Name Description
step1_gmx_image1 gmx_image Wrapper of the GROMACS trjconv module for correcting periodicity (image) from a given GROMACS compatible trajectory file.
step2_mdfeaturizer1 mdfeaturizer Obtain Molecular Dynamics features for PyTorch MDAE model training (dataset and statistics).
step3_build_model build_model Build a PyTorch autoencoder model (MDAE) from input statistics and hyperparameters.
step4_train_model train_model Train a PyTorch autoencoder model (MDAE) from an input dataset and initial model, optionally saving the trained model and training metrics.
step5_gmx_image2 gmx_image Wrapper of the GROMACS trjconv module for correcting periodicity (image) from a given GROMACS compatible trajectory file.
step6_mdfeaturizer2 mdfeaturizer Obtain Molecular Dynamics features for PyTorch MDAE model training (dataset and statistics).
step7_evaluate_model evaluate_model Evaluate a trained PyTorch autoencoder model (MDAE) on a dataset, computing loss and collecting model evaluation variables.
step8_make_ndx1 make_ndx Creates a GROMACS index file (NDX) from an input selection and an input GROMACS structure file.
step9_make_ndx2 make_ndx Creates a GROMACS index file (NDX) from an input selection and an input GROMACS structure file.
step10_gmx_rmsf1 gmx_rmsf Wrapper of the GROMACS module for calculating the Root Mean Square fluctuations (RMSf) of a given GROMACS compatible trajectory.
step11_gmx_rmsf2 gmx_rmsf Wrapper of the GROMACS module for calculating the Root Mean Square fluctuations (RMSf) of a given GROMACS compatible trajectory.
step12_feat2traj feat2traj Convert reconstructed molecular dynamics features (e.g. decoded coordinates) into an MD trajectory file using topology information from the stats file and/or a topology file.
step13_gmx_rmsf3 gmx_rmsf Wrapper of the GROMACS module for calculating the Root Mean Square fluctuations (RMSf) of a given GROMACS compatible trajectory.
step14_make_plumed make_plumed Generate a PLUMED input file, a features.dat file and a TorchScript (.ptc) model from a trained MDAE model and feature statistics, suitable for biased molecular dynamics simulations.

Outputs

ID Name Description Type
step1_gmx_image1_out1 output_traj_path Path to the output file.
  • File
step2_mdfeaturizer1_out1 output_dataset_pt_path Path to the output dataset model file.
  • File
step2_mdfeaturizer1_out2 output_stats_pt_path Path to the output model statistics file.
  • File
step3_build_model_out1 output_model_pth_path Path to save the model in .pth format.
  • File
step4_train_model_out1 output_model_pth_path Path to save the trained model (.pth). If omitted, the trained model is only available in memory.
  • File
step4_train_model_out2 output_metrics_npz_path Path save training metrics in compressed NumPy format (.npz).
  • File
step5_gmx_image2_out1 output_traj_path Path to the output file.
  • File
step6_mdfeaturizer2_out1 output_dataset_pt_path Path to the output dataset model file.
  • File
step6_mdfeaturizer2_out2 output_stats_pt_path Path to the output model statistics file.
  • File
step7_evaluate_model_out1 output_results_npz_path Path to the output evaluation results file (compressed NumPy archive).
  • File
step8_make_ndx1_out1 output_ndx_path Path to the output index NDX file.
  • File
step9_make_ndx2_out1 output_ndx_path Path to the output index NDX file.
  • File
step10_gmx_rmsf1_out1 output_xvg_path Path to the XVG output file.
  • File
step11_gmx_rmsf2_out1 output_xvg_path Path to the XVG output file.
  • File
step12_feat2traj_out1 output_traj_path Path to save the trajectory in xtc/pdb/dcd format.
  • File
step12_feat2traj_out2 output_top_path Path to save the output topology file (pdb). Used if trajectory format requires separate topology.
  • File
step13_gmx_rmsf3_out1 output_xvg_path Path to the XVG output file.
  • File
step14_make_plumed_out1 output_model_ptc_path Path to the output TorchScript model file (.ptc) for PLUMED's PYTORCH_MODEL action.
  • File
step14_make_plumed_out2 output_plumed_dat_path Path to the output PLUMED input file.
  • File
step14_make_plumed_out3 output_features_dat_path Path to the output features.dat file describing the CVs to PLUMED.
  • File

Version History

Version 1 (earliest) Created 27th Mar 2026 at 09:24 by Genís Bayarri

Initial commit


Frozen Version-1 77cf6a9
help Creators and Submitter
Creators
Submitter
Discussion Channel
BioExcel Workflows
Citation
Bayarri, G., & Hospital, A. (2026). CWL AutoEncoders for MD Analysis tutorial. WorkflowHub. https://doi.org/10.48546/WORKFLOWHUB.WORKFLOW.2144.1
License
Apache License 2.0 (Apache-2.0)
Activity

Views: 224   Downloads: 71

Created: 27th Mar 2026 at 09:24

help Tags

This item has not yet been tagged.

help Attributions

None

Total size: 27.8 MB
Powered by
 (v.1.17.3)
About WorkflowHub | Acknowledgements | Credits | Terms & Conditions | Privacy Policy | Cite us | Contact us
Copyright © 2008 - 2026 The University of Manchester and HITS gGmbH