Published March 4, 2026 | Version 1.0
Other Open

Graphormer-RT Fine-tuned Weights (HILIC)

  • 1. ROR icon University of Oregon
  • 2. ROR icon Pacific Northwest National Laboratory

Description

These weights are for Graphormer-RT and were fine-tuned from the original weights for improved HILIC performance. 

The original model was developed by Cailum M. K. Stienstra, Emir Nazdrajić, and W. Scott Hopkins, as described in: “From Reverse Phase Chromatography to HILIC: Graph Transformers Power Method-Independent Machine Learning of Retention Times,” published in Analytical Chemistry.

Code for training and inference using these weights is available in our repository: deep_metab.

Files

Files (213.6 MB)

Name Size Download all
md5:e942b8b80957ca29d0f3ed8c82ff9c1b
213.6 MB Download

Additional details

Software

Repository URL
https://github.com/kbcoulter/deep_metab
Programming language
Python , R , Shell , Nextflow
Development Status
Active