UniGeneX: Credible Gene Sets, Training Configuration and Output

Reproducibility Data for UniGeneX This repository contains the training inputs and pre-trained models to verify the installation and basic functionality of the UniGeneX without needing to retrain the models from scratch File Descriptions 1. Creating credible gene set Xenium_panel.npy / Xenium_Vannan_panel.npy: Gene panels from Xenium used to define and expand the credible gene set. default_census_vocab.json: Vocabulary file for credible gene set. pretrain_data_train_genes.npy: List of credible gene set. 2. Model training input config_templete_train.yaml: Template for setting up the training pipeline. config_templete_inference.yaml: Template for running inference using pre-trained models. Training_input.zip: Primary training datasets. Training_input_testdata.zip: Test data input files for UGE generation 3. Model training output Training_output.zip: Pre-trained model checkpoints, and UGE (training and testing data).