From: Prompt Engineering for Transformer-Based Chemical Similarity Search Identifies Structurally Distinct Functional Analogues (Under review)
Authors: Clayton W. Kosonocky, Aaron L. Feller, Claus O. Wilke, Andrew D. Ellington
Link to arXiv: https://arxiv.org/abs/2305.16330
Contained herein are scripts for the described Chemical Semantic Search (CheSS), alternative canonicalization prompt engineering method, and various analyses of the search results.
Due to the constraints on GitHub filesize:
- The database file "pubchem-9999810-512-token-compatible.txt" is not available
- The search results only include the top 1,000 results, rather than the entire database
If you wish to obtain copies of these files, please contact [email protected]. We may host these files on a separate storage platform, in which case we will post a link here.
@misc{kosonocky2023prompt,
title={Prompt Engineering for Transformer-based Chemical Similarity Search Identifies Structurally Distinct Functional Analogues},
author={Clayton W. Kosonocky and Aaron L. Feller and Claus O. Wilke and Andrew D. Ellington},
year={2023},
eprint={2305.16330},
archivePrefix={arXiv},
primaryClass={physics.chem-ph}
}