SAMPL5 Host-Guest Binding Free Energies with Water Analysis

Valerio Rizzi from ETH Zurich combined machine learning with physical intuition to build a water-describing collective variable for atomistic simulations. This variable was used to study host-guest systems from the SAMPL5 challenge, obtaining highly accurate binding free energies and good agreement with experiments. The dataset likely contains simulation results and analyses detailing the role of water during the binding process.

Use Cases

• Benchmarking enhanced sampling techniques based on the described non-local water collective variable.
• Analyzing the role of water in ligand binding processes based on the detailed study mentioned.
• Training or validating machine learning models for molecular simulation based on the SAMPL5 challenge systems used.
• Studying kinetic bottlenecks in free energy landscapes based on the simulation challenges described.

Strengths

• Results show good agreement with experimental binding free energies for the SAMPL5 challenge systems.
• The method combines machine learning with physical intuition, as described in the abstract.
• The study focuses on a known challenge area: capturing the complex behavior of water in binding.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

paperswithcode, author Valerio Rizzi, organization ETH Zurich

Collection Method

Atomistic computer simulations using a machine-learned collective variable, as described in the abstract.