Back

QuantCell: machine learning based cell annotation from qualitative and quantitative imaging profiles

Boohar, W. R.; Wang, B.; Thomas, Z.; Nogalska, A.; Lu, R.

2026-02-17 bioinformatics
10.64898/2026.02.15.706033 bioRxiv
Show abstract

Recent advances in spatial omics enable high-resolution, multiplexed imaging of RNA and protein expression, but cell annotation remains challenging, particularly in complex tissues with numerous markers or rare cell types. Here, we present QuantCell, a machine learning framework that leverages quantitative imaging data to improve annotation derived from qualitative profiles. QuantCell evaluates multiple models and applies a user-defined false discovery rate to ensure high-confidence annotation. Using PhenoCycler imaging of mouse bone marrow, QuantCell increased annotated cells from 33.1% to 90.2% at 5% FDR, achieving 96.5% accuracy. QuantCell supports diverse imaging platforms and robustly detects rare cell populations.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.