search

stuart-lab / signac

R toolkit for the analysis of single-cell chromatin data
https://stuartlab.org/signac/
Other
328 stars 88 forks source link

Custom genome #1

Closed sameet closed 5 years ago

sameet commented 5 years ago

Hi,

How do I go about using signac with custom genome, for which the EnsDb, and BSgenome is not yet available. Do I need to start with those two first?

Sameet

timoast commented 5 years ago

These are not necessary for analysis with Signac and you'll be able to do most things without it.

One thing we use the EnsDb for is supplying the gene coordinates in order to create a gene activity matrix (with the FeatureMatrix function), but this could be done in other ways (reading a BED or GFF file for example). You just need to supply a GRanges object containing the genomic intervals you want to quantify for each gene.

We also use the EnsDb annotations for plotting the gene models alongside coverage tracks, but this is optional and you can plot the coverage tracks with CoveragePlot without supplying an EnsDb annotation.

The BSgenome is needed for motif analysis, which is still under development.

sameet commented 5 years ago

Is it possible to use the "gene-activity" object that was created for say CICERO with signac.

timoast commented 5 years ago

Yes, you can add any matrix as a new assay to the object, so you could create a gene activity matrix with Cicero and then add it to the Seurat object the same way we do in the vignette, ie:

object[["RNA"]] <- CreateAssayObject(counts = cicero_data)
sameet commented 5 years ago

Wonderful!! I think this will solve the problem that I currently have. I will try this and get back to you with how it goes.
Thank you.

timoast commented 5 years ago

Great, good luck!

kyrenya commented 4 years ago

Hi there! Have now the same question, but specifically for Motif analysis. Is it possible to do it without a BSGenome? Thanks in advance!

timoast commented 4 years ago

You can find enriched motifs without a BSgenome. The only requirement is a peak x motif matrix describing which peaks contain which motifs. However, in practice, it would be difficult to generate that matrix without a BSgenome object. See here for documentation for creating a BSgenome object: https://www.bioconductor.org/packages/release/bioc/vignettes/BSgenome/inst/doc/BSgenomeForge.pdf

kyrenya commented 4 years ago

Ok, I imagined that would be the case... Thanks a lot for your fast reply!