fidelram
commented
5 years ago I don't think that what you want is possible.
sc.pl.tracksplot plots data in a way that resemble a genome browser track but ist not because it does not understand coordinates. It simple groups the cells by the given groupby condition and then plots the value of each gene in a separate track. The y value is the gene expression (or in your case the ATAC-seq value). The x coordinate, simple puts all cells one after the other without any ordering.
On Sat, Jan 26, 2019 at 12:55 AM manarai notifications@github.com wrote:
Hi,
Thanks for this amazing package.
I have been playing with scanpy on scATACSeq data generated from 10x. And in comparison to the cellranger analysis, I think analysis scanpy does pretty descent job and adds more possibilities. I would like to displays some peaks that are highly present if some clusters using the genome browser which scanpy seem to be able to do "I think" ( as shown below). Is it possible to the same thing but with the peak averaged for all cells within the same cluster?
import matplotlib.pyplot as plt genes =['chr15:101708546_101718131','chr11:117961932_117970696', 'chr19:5821847_5852441','chr15:101422873_101429606', 'chr17:39842811_39849028','chr13:6108971_6109684'] sc.pl.tracksplot(adata,genes,groupby='louvain', figsize=[40,50])
[image: atacseq] https://user-images.githubusercontent.com/39877296/51778958-d7e4c700-2147-11e9-88cd-78f3e100c75f.png
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/theislab/scanpy/issues/447, or mute the thread https://github.com/notifications/unsubscribe-auth/AEu_1VUU1fUPbHKqadt-yXNKrA4amCXQks5vG5lngaJpZM4aT2VQ .
manarai
Hi,
Thanks for this amazing package.
I have been playing with scanpy on scATACSeq data generated from 10x. And in comparison to the cellranger analysis, I think analysis scanpy does pretty descent job and adds more possibilities. I would like to displays some peaks that are highly present if some clusters using the genome browser which scanpy seem to be able to do "I think" ( as shown below). Is it possible to the same thing but with the peak averaged/concatinated for all cells within the same cluster?
import matplotlib.pyplot as plt genes =['chr15:101708546_101718131','chr11:117961932_117970696', 'chr19:5821847_5852441','chr15:101422873_101429606', 'chr17:39842811_39849028','chr13:6108971_6109684'] sc.pl.tracksplot(adata,genes,groupby='louvain', figsize=[40,50])