RT Journal Article
SR Electronic
T1 Northstar enables automatic classification of known and novel cell types from tumor samples
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 820928
DO 10.1101/820928
A1 Fabio Zanini
A1 Bojk A. Berghuis
A1 Robert C. Jones
A1 Benedetta Nicolis di Robilant
A1 Rachel Yuan Nong
A1 Jeffrey Norton
A1 Michael F. Clarke
A1 Stephen R. Quake
YR 2020
UL http://biorxiv.org/content/early/2020/04/25/820928.abstract
AB Single cell transcriptomics is revolutionising our understanding of tissue and disease heterogeneity, yet cell type identificationl remains a partially manual task. Published algorithms for automatic cell annotation are limited to known cell types and fail to capture novel populations, especially cancer cells. We developed northstar, a computational approach to classify thousands of cells based on published data within seconds while simultaneously identifying and highlighting new cell states such as malignancies. We tested northstar on human glioblastoma and melanoma and obtained high accuracy and robustness. We collected eleven pancreatic tumors and identified three shared and five private neoplastic cell populations, offering insight into the origins of neuroendocrine and exocrine tumors. northstar is a useful tool to assign known and novel cell type and states in the age of cell atlases.Competing Interest StatementThe authors have declared no competing interest.