emarro/hg38_512_overlap_256
收藏官方服务：
资源简介：
--- configs: - config_name: default data_files: - split: train path: "train/*.arrow" - split: validation path: "validation/*.arrow" --- Processed version of hg38 with a seq length of 512bps and an overlap of 256 bps. Chr8 is left out of training set. Sequences with more than 25% being "N"s are filtered out.
提供机构：
emarro
