Researchers at Peking University and other centers present findings from a single-cell RNA sequencing analysis of colorectal cancer (CRC) organoids grown in condition medium or a chemical-defined medium. The team profiled more than 6,200 individual cells from CRC organoid, primary tumor, or matched normal samples representing seven CRC patients and analyzed the transcriptome data in conjunction with corresponding exome sequence, whole-genome sequence, targeted sequencing, or bisulfite sequencing-based DNA methylation data. Among other findings, the authors note that "organoids derived from tumor tissues faithfully recapitulate the main gene expression signatures of cancer cells in vivo," while normal tissue organoids "exhibited some tumor-like features at the whole transcriptome level" despite normal point mutation, copy number, and other genomic features.
A team from the Weizmann Institute of Science describes an algorithm for analyzing single-cell RNA-seq datasets representing a few cells to millions of cells — an approach the group applied to scRNA-seq data for up to 380,000 human bone marrow cells and to single-cell transcriptome profiles representing some 1.8 million to mouse embryonic cells. The scalable analytical method, known as Metacell-2, relies on a "recursive divide-and-conquer algorithm allowing efficient decomposition of scRNA-seq datasets of any size into small and cohesive groups of cells called metacells," the researchers explain, noting that the current version of the algorithm "improves outlier cell detection and rare cell type identification."
Investigators at the University Health Network, the University of Toronto, and the Vector Institute outline a machine learning-based approach for analyzing scRNA-seq profiles across datasets. The team calls its "One Cell at a Time" (OCAT) approach a "fast and memory-efficient machine learning-based method that does not require explicit batch effect removal in integrating multiple scRNA-seq datasets" and note that the freely available OCAT package makes it possible to do downstream analyses on the integrated scRNA-seq datasets, including cell type inference and differential gene selection. "OCAT effectively facilitates a variety of downstream analyses with important biological implications," the authors write, noting that the method "can undertake challenging tasks such as differential gene expression analysis, trajectory inference, pseudotime inference, and cell type inference."