Reproduce “ChromFound: Towards A Universal Foundation Model for Single-Cell Chromatin Accessibility Data” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yifeng Jiao, Yuchen Liu, Yu Zhang, Xin Guo, Yushuai Wu, Chen Jiang, Jiyang Li, Hongwei Zhang and 4 more
ChromFound represents single-cell chromatin accessibility profiles using genomic coordinates and continuous accessibility for each open region, combining local attention with long-range context in a pretrained model. Trained on 1.97 million human-cell profiles, it achieved an adjusted Rand index of 0.6890 on Morabito130K cortex batch 1 without target-dataset training, versus 0.5460 for SCALE; this result does not establish universal performance.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.