HBV splicing project: transcriptomes, genomic references, and AI models for splice site prediction

This archive supports the study "Decoding the interconnected splicing patterns of hepatitis B virus and host using large language and deep learning models" (Lim CS, Brown CM. 2025. BioRxiv. https://doi.org/10.1101/2025.07.28.667110). It provides the following contents needed to reconstitute the full project structure of the HBV_splicing_paper_2025 GitHub repository. Contents data/: Processed HBV transcriptomes from liver biopsies and cultured cells ref/: Reference genomes and annotations for human and HBV src/: Source code, pretrained models, and checkpoints for: SpliceBERT (LLM framework) OpenSpliceAI (convolutional neural network) IPSA (splicing metric calculation) These directories are intended to be unzipped into the root of the GitHub repository to reproduce the complete project structure. Metadata for these files is maintained in the GitHub repository. License Summary This archive includes code and AI models under the following licenses: src/SpliceBERT/: BSD 3-Clause src/OpenSpliceAI/: GNU GPLv3 src/IPSA/: GNU GPLv2