SpaRED benchmark: Enhancing Gene Expression Prediction from Histology Images with Spatial Transcriptomics Completion

TL;DR

This paper introduces a comprehensive benchmark dataset and a transformer-based method for predicting gene expression from histology images, improving accuracy and enabling fair comparisons across studies.

Contribution

It provides the largest curated database for gene expression prediction and proposes a novel transformer-based completion technique for missing data imputation.

Findings

8.6-fold increase in dataset size

Transformer-based completion significantly improves prediction accuracy

Establishes the most comprehensive benchmark to date

Abstract

Spatial Transcriptomics is a novel technology that aligns histology images with spatially resolved gene expression profiles. Although groundbreaking, it struggles with gene capture yielding high corruption in acquired data. Given potential applications, recent efforts have focused on predicting transcriptomic profiles solely from histology images. However, differences in databases, preprocessing techniques, and training hyperparameters hinder a fair comparison between methods. To address these challenges, we present a systematically curated and processed database collected from 26 public sources, representing an 8.6-fold increase compared to previous works. Additionally, we propose a state-of-the-art transformer based completion technique for inferring missing gene expression, which significantly boosts the performance of transcriptomic profile predictions across all datasets. Altogether, our contributions constitute the most comprehensive benchmark of gene expression prediction from histology images to date and a stepping stone for future research on spatial transcriptomics.