search

GreenleafLab / ArchR

ArchR : Analysis of Regulatory Chromatin in R (www.ArchRProject.com)
MIT License
384 stars 137 forks source link

Getting IDs of Cells in Clusters #480

Closed spietrzak closed 3 years ago

spietrzak commented 3 years ago

I have used ArchR to perform clustering of cells from scATAC-seq data obtained from the 10X Genomics platform. I was wondering how I might be able to retrieve the IDs (barcodes) of the cells that make up each of the clusters. For example, what are the barcodes of all of the cells that make up Cluster 1, etc. Any information you have on this would be great.

rcorces commented 3 years ago

Look in cellColData https://www.archrproject.com/bookdown/manipulating-an-archrproject.html

Please close if this answers your question.

spietrzak commented 3 years ago

I have a follow up question. I obtained the cluster, sample, and barcode data for the cells in two separate analyses I performed in ArchR (in 2 different R-projects). Some of the samples that were used in these analyses were used in both. However, when I overlap the barcodes associated with a particular sample from these two different projects, they are not exactly the same. The number of barcodes is the same, but there is a small handful of barcodes that are unique. I believe that the barcodes should be the same since it is the same fragments file being input into both projects, so I thought it was strange that there was a small group (30 barcodes out of ~3,000) that were different between them. I am wondering if you have an explanation for this discrepancy.

rcorces commented 3 years ago

Unless you used slightly different parameters for createArrowFiles(), I would not expect this to be the case. Can you look at the barcodes that are different and see if there is anything particular about those cells that might have caused this? Have you tried creating Arrow files multiple times and making sure that the barcodes are identical?

rcorces commented 3 years ago

Closing due to inactivity. Feel free to comment again here and I will re-open if you still need help.