Reproduce “Spatially Informed and Direction Sensitive Representation Learning for 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

Azim Dehghani Amirabad, Ali Khodabandeh Yalabadi, Artem Moskalev, Yun-Ching Chen, Mehdi Pirooznia, Scott Oloff, Rui Liao

Tessera combines each cell’s gene-expression profile with nearby cells’ relative displacement, allowing attention to reflect both distance and direction. It trains on masked-cell expression reconstruction. Across nine held-out sections, RoPE achieved Pearson r of 0.4311, compared with 0.2575 without positional information. These results support imputation and representation evaluation, not therapeutic or clinical outcomes.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.