Bioinformatics Seminar
Time: 11AM
Venue: Davis Auditorium and Online
5 March 2024
Φ-space: continuous phenotyping of single cells
Kim-Anh Le CaoMIG/The University of Melbourne
Cell type annotation is a key step in single-cell and spatial omics analysis. Computational approaches typically utilise well-curated and annotated transcriptomics, based on either bulk or single-cell RNA-seq datasets as the reference to annotate query data. With the advancement of sequencing technologies, new query datasets are being generated using radically different omics platforms compared to the reference. We developed Φ-space (Phi-space), a computational framework for the continuous phenotyping of single-cell multiomics and spatial transcriptomics data. Φ-space uses Projection to Latent Structures regression to achieve robust and continuous annotation of query cells and spatial pixels. We show that continuous rather than discrete annotation is more useful for the discovery of new cell states and spatial patterns in several case studies.