GNNePCSAFT MCP Server

Example Use Cases

• 1

  Estimate PC‑SAFT parameters for novel molecules using a GNN and run PC‑SAFT property predictions without experimental inputs.

• 2

  Compute thermodynamic properties (density, vapor pressure, enthalpy of vaporization, critical points) for pure components and mixtures programmatically.

• 3

  Automatic molecular data retrieval from PubChem and integration of those data into thermodynamic calculation pipelines.

Description

This MCP server implements the Model Context Protocol to allow LLMs and external clients to request PC‑SAFT parameter estimation and thermodynamic property calculations using GNNePCSAFT (GNN-based parameter prediction) and FeOS for PC‑SAFT computations. It can automatically fetch molecular data from PubChem, handle pure components and mixtures, and return properties such as density, vapor pressure, and enthalpy of vaporization. The server is intended to enable thermodynamics-aware agents and automate property estimation for molecules with little or no experimental data.

Capabilities(10 total)