Tutorials

Getting Started

These tutorials describe the basics of mlcolvar through an example workflow. In addition, instructions on how to create datasets and customize the training are also given.

• Getting Started
  • Training your first collective variable
  • Creating datasets
  • Customize training

Collective variables

The following tutorials cover (some of) the CVs implemented in the library.

• Methods for CVs optimization
  • AutoEncoding collective variables
  • Deep-LDA: Neural network-based discriminant CVs
  • Deep-TDA: Deep Targeted Discriminant Analysis
  • Deep-TICA: Neural network basis functions for TICA
  • Learning the committor

Customizing CVs

How to customize the CVs, either using the multitask framework or by creating new ones based on the existing ones. Moreover, we also discuss how to add pre or post processing layers.

• Customizing CVs
  • Multi-task framework
  • Implementing a new CV from scratch
  • Implementing new CVs via subclassing
  • Pre/post processing
  • Transform module of mlcolvar

Interpreting CVs

Some examples of how to interpret the CVs, using sensitivity analysis or sparse linear models.

• Interpreting CVs
  • Sensitivity analysis
  • Sparse linear models

Utils

Tutorials on how to use some of the utils functions provided.

• Utils
  • Utils: Compute and plot free energy surface and free energy differences