Leveraging Large Language Models for Decision Support in Precision Oncology

This repository contains all scripts and data to reproduce the experiments done in our paper "Leveraging Large Language Models for Decision Support in Precision Oncology".

Here's an overview about the contents of the repository:

• The conversation_logs folder contains the full chat logs with each of our examined LLM (ChatGPT 3.5, Perplexity.ai, BioMed LM and Galactica).

• The data folder contains all the information about the ten fictional patient cases, the MTB survey results and all the cleaned up and pre-processed output for each LLM.

• The figures and tables folders contain the results from our analyses. To regenerate the analysis results by yourself, re-run the R-script analysis.R.