Gruffi removes stressed cells from brain organoid datasets

Organoids enable disease modeling in complex and structured human tissue, in vitro. Like most 3D cultures, they lack sufficient oxygen supply, leading to cellular stress. These negative effects are particularly prominent in complex models, like brain organoids, where they can prevent proper lineage commitment. Here, we analyze brain organoid and fetal single cell RNA sequencing (scRNAseq) data from published and new datasets totaling about 190,000 cells. We describe a unique stress signature found in all organoid samples, but not in fetal samples. We demonstrate that cell stress is limited to a defined organoid cell population, and present Gruffi, an algorithm that uses granular functional filtering to identify and remove stressed cells from any organoid scRNAseq dataset in an unbiased manner and validate our findings on six further datasets from different organoid protocols and early brains. Our data show that adverse effects of cell stress can be corrected by bioinformatic analysis, improving developmental trajectories and resemblance to fetal data. Overall design: Generation of new and re-analysis of published organoid scRNA seq datasets