| layout | title |
|---|---|
home |
Welcome to Celligrate! |
Celligrate is a project for cell type characterization and integration from single cell RNA-sequencing (scRNA-seq) data. The backbone of Celligrate consists of two carefully-designed and extensively-validated computational algorithms: NS-Forest and FR-Match. NS-Forest is a random forest machine learning algorithm for cell type marker gene identification. FR-Match is a topological graph theory-based statistical learning algorithm for cell type matching. Celligrate also introduces a notion of “cell type barcode” for insightful visualization of cell type expression data. The use of Celligrate extends the utility of the upstream scRNA-seq analysis pipelines to downstream use cases, and ultimately accelerates the growth of knowledge about cell types by pooling results from individual studies.
See Methods for more details.
{:height="150px" width="125px"}
{:height="150px" width="125px"}
{:height="150px" width="125px"}
- Please find NS-Forest v4.0 tutorial on readthedocs
- NS-Forest v3.9 tutorial
- Tutorial 1: FR-Match data object + matching across overlapping neuroanatomical regions (Layer 1 and full MTG)
- Tutorial 2: Matching across non-overlapping neuroanatomical regions (M1 and MTG)
- Tutorial 3: Matching mouse M1 cell subclasses across different sample types (scRNAseq and snRNAseq)
- Tutorial 4: Matching human M1 cell subclasses across different platforms (SMART-seq and 10X)
- Tutorial 5: Cell type calling for spatial transcriptomics data (smFISH and scRNAseq)
- Tutorial 6: Notebook for challange 1 of Allen Institute Data challenge: Mapping of Cell Type Data
For inquiries of demo data, please contact [email protected].