MorPhiC Hackathon San-Fransisco 2025 TF-target data

This repository contains the data related to the MorPhiC Hackathon San-Fransisco 2025 task 1:<b>TF - target inference (from single-cell data).</b>The h5ad file is a file that can be loaded with Scanpy and contains <i>3</i>5,799 wild-type (WT) cells and 80,544 TF knockout (KO) cells spanning nine cell types involved in the differentiation of endocrine cell types from precursors within the pancreas and 27 TFs.The cell metadatas are available directly within the anndata (anndata.obs) and in the meta.all.tsv file included in the dataset.The index_H1.main_type.tsv indicates the WT cells used for the analysis<b>TF-target Gene Regulation Networks (GRNs)</b>The GRN files contain the inferred GRNs for 4 approaches and each of the 9 endocrine cell types: GRNboost2, SCENIC, CellOracle, and Inferelator.In addition, GRNs inferred only from biomedical corpuses (LLM) or databases are available in the grn.llm file.<b>TF-&gt;target small scale interaction formatted for LLM training</b>llm_target_question.tar.gz contains 3 json files (training, test, validation) used to train a LLM model.<br>