Reproduce “COAST: Context-Aware Differential Learning for Gene Expression Prediction in Spatial Transcriptomics” 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
Keunho Byeon, Sunhong Park, Jeewoo Lim, Jin Tae Kwak
COAST predicts gene expression from H&E images by combining target spots with local and slide-wide context. It trains one head to estimate absolute expression and another to predict signed expression differences between target and context spots. Across seven SpaRED datasets, COAST reports best aggregate results on five metrics, with mean PCC of 0.4262 ± 0.08, though not in every comparison.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.