data.gov.au
Resradar
Published by CSIRO Data Access Portal
About this dataset
This repository contains the code and data-processing pipelines for the paper "How similar are protein crystallisation conditions? A Data-Driven Deep Learning Approach to Quantifying Similarity". It learns a similarity metric for protein crystallisation reservoir conditions. The pipeline preprocesses screening data, calculates baseline chemical distances, and trains a transformer model to embed reservoir conditions. The model positions chemically and experimentally similar conditions closer together in the embedding space.