Summary
BioKERN is a multimodal spatial representation-learning framework for histology-to-transcriptomics mapping that treats biological structure as an explicit, learnable inductive bias. Instead of emphasizing instance-level cross-modal matching between paired spots, BioKERN constructs a training-time biological kernel by combining transcriptomic similarity and spatial proximity, and uses it to provide graded neighborhood supervision while regularizing embedding geometry. This encourages representations that preserve biological neighborhood structure rather than only exact cross-modal correspondences. Evaluation uses a fixed, model-independent biological neighborhood definition shared by all methods, enabling fair comparison. Across Mouse Brain Visium and Human Liver GSE240429 datasets, BioKERN consistently improves biological neighborhood retrieval in both single-scale and multi-scale settings. The paper (arXiv:2608.24823) is authored by Seungik Cho and Betul Orcan-Ekmekci in the machine learning domain.
Overview
Research area: ML
Authors: Seungik Cho, Betul Orcan-Ekmekci
Published: 2026-08-25
arXiv: 2608.24823
Abstract
Spatially resolved biology requires representations that preserve biological neighborhood structure rather than only exact cross-modal correspondences. Existing histology--transcriptomics objectives can emphasize instance-level matching even when non-paired spots share molecular or spatial context.
The authors introduce BioKERN, a multimodal spatial representation-learning framework that incorporates biological structure as an explicit, learnable inductive bias. BioKERN constructs a training-time biological kernel by combining transcriptomic similarity and spatial proximity, then uses it to provide graded neighborhood supervision and regularize embedding geometry.
Evaluation uses a fixed, model-independent biological neighborhood definition shared by all methods. Across Mouse Brain Visium and Human Liver GSE240429 datasets, BioKERN consistently improves biological neighborhood retrieval in single-scale and multi-scale settings.
---
*Auto-collected on 2026-08-27.*
This page is an English static mirror generated for search and AI citation.
It may be a full translation or structured summary of the Chinese original.
Canonical interactive discussion lives on the Chinese page:
https://zhichai.net/topic/178634095