Newest papers

Spatially Informed and Direction Sensitive Representation Learning for Spatial Transcriptomics

Azim Dehghani Amirabad, Ali Khodabandeh Yalabadi, Artem Moskalev, Yunching Chen, Mehdi Pirooznia, Scott Oloff, Rui Liao · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.28579 · cited by 0 works (OpenAlex) · open-access PDF

Portuguese

Os estados do contexto celular são moldados por programas de expressão de genes intrínsecos e pelo microambiente espacial em que as células residem. A transcriptomia espacial torna esse contexto mensurável, mas o aprendizado de representação eficaz requer a captura de ambos os sinais moleculares informativos espacialmente e da organização intercelular. As abordagens existentes abordam esses desafios apenas parcialmente: a seleção de genes convencionais pode ignorar sinais estruturados espacialmente, enquanto os métodos de codificação de relações células-celulares muitas vezes reduzem o microambiente de uma célula à proximidade sozinho, descartando a direção, ou dependem de coordenadas absolutas ligadas a um quadro de tecido arbitrário. Consequentemente, as células com distâncias vizinhas semelhantes, mas organização

Machine translation (Meta M2M100); the English original follows.

Abstract

Cellular states are shaped by both intrinsic gene expression programs and the spatial microenvironment in which cells reside. Spatial transcriptomics makes this context measurable, but effective representation learning requires capturing both spatially informative molecular signals and intercellular organization. Existing approaches address these challenges only partially: conventional gene selection can overlook spatially structured signals, while methods for encoding cell-cell relationships often either reduce a cell's microenvironment to proximity alone, discarding direction, or rely on absolute coordinates tied to an arbitrary tissue frame. Consequently, cells with similar neighbor distances but distinct spatial organization may receive overly similar representations, despite experiencing different biological contexts that shape their cellular states. We introduce Tessera, a spatial representation learning framework that addresses both molecular selection and cellular context. Tessera combines spatially informed gene selection with a geometric inductive bias that contextualizes each cell within its local tissue environment. By modeling relative cell-cell relationships, Tessera is translation-invariant while preserving direction, allowing neighboring cells to contribute differently according to their spatial context and cellular state. Across held-out tissues and technologies, Tessera outperforms existing spatial foundation models in representation quality and generalizes zero-shot to contexts up to $16\times$ larger than those seen during training.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Representation (politics), Context (archaeology), Computer science, Selection (genetic algorithm), Artificial intelligence, Spatial contextual awareness

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange