Integrate scRNA data with TF binding site prediction analysis. These scripts are still under development and are currently being used only internally by the SchulzLab group members. A copy of all these files can be found at /home/skarunanithi/identifyEnrichedTFsWithscRNAdata/ in our servers.
The workflow is as follows.
- Step-1: For a gene of interest get the REMs from the EpiRegioDB as a CSV file
- Step-2: We run the scripts to identify enriched TFs in the TF Binding Sites which can be found in the REMs using PASTAA and FIMO (created by Nina). This step gives a list of enriched TFs and their p-values. The files associated with this step are located in src, PASTAA_Fimo_analysis, identifyEnrichedTFs, and identifyTFBindingSites.
- Step-3: We run MetaCellaR (created by Fatemeh) to create Metacells from the available single cell data. The current version of MetaCellaR can be found at https://github.com/SchulzLab/MetaCellaR.
- Step-4: Using the Metacells identified in Step-3, we calculate the RNA expression correlation of your gene of interest against all genes, and their p-values. You can subset this step to include only a particular cell type of your interest. If the interested cell type is unavailable, all available cell types will be used.
- Step-5: We calculate a meta p-value (uses the metaseqR package) by integrating the p-values from Step-2 and Step-4. Please note that this step will report the meta p-values for only the enriched TFs identified in Step-2.
bash ~/identifyEnrichedTFsWithscRNAdata/scripts/Pipeline.sh -o=./testNOX4 -e=NOX4_REMs.csv -mcr=~/identifyEnrichedTFsWithscRNAdata/MetaCellaR -scf=./droplet_Liver_seurat_tiss.Robj -r='data' -c='meta.data$cell_ontology_class' -u=T -g=NOX4 -cell='hepatocyte'
- PASTAA (available in scripts folder)
- FIMO (available in scripts folder)
- R packages (for MetaCellaR and post-processing): metaseqR, Seurat, cluster, knn.covertree, irlba, ggplot2, Rtsne, umap
| Short | Long | Description |
|---|---|---|
| -h | --help | Display Help commands |
| -o | --outloc | Path to the output directory (will be created if not available |
| -e | --epiregio | Epiregio ouptut for the gene of your interest downloaded as a CSV file |
| -gseq | --genome_file | genome sequence file in FASTA format (Default: ~/identifyEnrichedTFsWithscRNAdata/scripts/hg38.fa) |
| -tfw | --tfbinding_workflow_loc | Path to the TRAP/FIMO workflow created by Nina (Default: ~/identifyEnrichedTFsWithscRNAdata/scripts) |
| -tfw_scr | --tfbinding_scRNA_workflow_loc | Path to scRNA integration pipeline created by Siva, which has additional processing scripts (Default: ~/identifyEnrichedTFsWithscRNAdata/scripts) |
| -mcr | --metacellar_loc | Path to MetaCellaR created by Fatemeh (Default: ~/identifyEnrichedTFsWithscRNAdata/MetaCellaR) |
| -scf | --scRNA_file | Name of the scRNA file containing a Seurat object |
| -r | --scRNA_slot | Name of the RNA expression data slot in the Seurat object |
| -c | --scRNA_cell_annot_slot | Name of the cell type annotation data slot in the Seurat object |
| -u | --run_umap | Boolean T or F, whether to run UMAP as part of MetaCellaR (Default: T) |
| -g | --gene_name | Name of the gene of interest (Eg. NOX4) |
| -cell | --cell_name | Name of the interested cell type (if any; Default: None) |