Scripts for Interpretable Machine Learning to Understand Wildfire Toxicity: Bridging Chemicals and Multi-Omics
The aim of this project was to develop and apply code for symbolic regression (SR)-based analyses to bridge chemical exposure data and transcriptomic responses to wildfire smoke. This repository contains all the code, data inputs, intermediate files, model outputs, and figures used to support the analyses described in the project.
Includes raw input data used for model training and evaluation. These datasets include chemical exposure and transcriptomic response profiles.
Contains all scripts used for data preprocessing, model training, evaluation, and figure generation.
Stores processed datasets and intermediate files generated during the data cleaning, transformation, and analysis pipeline.
Contains output files from trained SR models, including models selected for interpretation.
Includes all generated figure images as well as supporting plots used for interpretation.