Disease related cell type analysis to decode cell type effect and underlying regulatory mechanisms across human disease
-
Formatting data from different sources
-
Preprocessing data to get the following ouputs:
- scRNA-seq: seurat object with cell annotations in metadata
- scATAC-seq: seurat or ArchR object with cell annotations in metadata
-
Calculate the cell specific genomic intervals by scATAC-seq
3.1 Call Cell specific open chromatin regions were used as input for LDSC
3.2 Using LDSC to identify disease related cell type marker genes.
3.3 Downstream analysis:
- (I) Overlap disease related SNPs and cell-type specific variable CREs. In disease related samples, we suggest these SNPs open chromatin to cause disease. In normal sample, we suggest these SNPSs close chromatin to cause disease.
- (II) (Optional) Identify cell type specific variable genes that are related SNPs by overlapping the gene's TSS (Transcription Start Site) region with the SNPs.
- (III) Perform a Transcription Factor (TF) targeting analysis in the CREs identified in part I (significant motif related tf as tf, region annotated gene as target). This will help us further understand how the TFs might play a role in regulating specific genes in respective cell types.
- (IV) When scATAC-seq data is available, integrating Pearson correlation and TF-target result of part IIII to reconstruct cell type-specific gene regulatory network. When scATAC-seq data is not available, using SCENIC to reconstruct cell type-specific diseased related gene regulatory network.
We also have a step-by-step tutorials to preprocess single cell multiomics data in reproduce_case folder
DRCTDB is a fully open-source database, where users can download the entire project and run locally at db-core folder