szawan/llm_xai_biomarkers

This repository contains the code and files used in the manuscript "Using LLM Models and Explainable ML to Analyse Biomarkers at Single Cell Level"

In addition to the code and files in the repository, the following link should be used:

https://drive.google.com/drive/folders/1jyl0esSY8VYMQ1Ke7P-E-VMidR022awM?usp=drive_link

The link allows one to download the following files:

• The fully trained scBERT model trained on the Gut Cell Atlas
• The four UC preprocessed datasets used for constructing the QLattice models, which can be annotated using the .txt files in the scBERT_Folder/Predictions folder
• The Zheng68k dataset used for scBERT and the preprocessed scRNA-seq dataset corresponding to the Gut Cell Atlas