HDMA ChromBPNet models - Part 2

Trained ChromBPNet models (including scaled bias model, residual accessibility model, and full model), for five folds for each cell type, for HDMA (Liu*, Jessa*, Kim*, Ng* et al, bioRxiv 2025). These are gzipped tar archives which can be extracted with `tar -xvzf <file.gz>`. A detailed description of the main data types deposited on Zenodo can be found here.已训练的ChromBPNet模型（包含缩放偏置模型、残差可及性模型与完整模型），对应HDMA研究中每种细胞类型的5折拆分（Liu*、Jessa*、Kim*、Ng*等，bioRxiv 2025）。本数据集为gzip压缩的tar归档文件，可通过`tar -xvzf <file.gz>`命令完成解压。上传至Zenodo的核心数据类型详细说明可参见此处。