search

stuart-lab / signac

R toolkit for the analysis of single-cell chromatin data
https://stuartlab.org/signac/
Other
318 stars 86 forks source link

Z-scores of cis-regulatory elements and TF motif #379

Closed Ping-lin14 closed 3 years ago

Ping-lin14 commented 3 years ago

Hi timoast

I want to know if there is any way in Signac to implement this paper (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7299161/#SD2), Fig 2-c and Fig 5-g Analysis method.

Thanks.

timoast commented 3 years ago

2C is a heatmap of z-scores, which you could create using ScaleData and DoHeatmap.

5G is a heatmap of averaged accessibility over pseudotime, with accessibility at each pseudotime point shown as the percentage of maximum accessibility. We don't have functionality to make smoothed heatmaps currently in Seurat/Signac and it looks like Monocle has dropped that functionality also. However, it should be fairly straightforward to do outside of Seurat/Signac by extracting the data from the object (d <- GetAssayData(...)), divide each row by the row maximum (d <- d / apply(d, 1, max)) , then create a heatmap.

Ping-lin14 commented 3 years ago

Thank you very much for your suggestion.

For the heat map of z scores, I got this result(https://ibb.co/fqYBpJB). I want to convert the open regions of chromatin into gene names, but using Closest Feature() will lose logFC information. Is there any other method?

timoast commented 3 years ago

I don't quite understand what you're wanting to do, there's no fold-change information involved in the z-score or needed for the heatmap. ClosestFeature() will give you the gene annotation closest to each region, but I would caution against interpreting open chromatin regions as simply regulating the closest gene.