How does human subjectivity affect the reliability of single-cell RNA-seq cell-type annotation?
Under review at Cell Systems
This study examines how analyst subjectivity affects scRNA-seq cell-type annotation, finding substantial variability across 37 participants and a 58% higher false positive rate when analysts expected specific rare cell types. I contributed statistical analysis of parameter tuning effects, characterization of analyst background influences on accuracy, and data preprocessing; under review at Cell Systems.