PeakSeuratFromTransfer or NewPeakSeurat Error about sierra

[rj-patrick]

Hi Ashley,

My guess is there is problem with your annotations table - would you be able to show a few lines from the table with head(peak_annotations)?

Cheers,
Ralph

[aralbright]

Sure!

1 | Act87E:3R:13427630-13428086:1
2 | Mps1:3R:17673232-17673828:1
3 | Ama:3R:6764596-6765261:1
4 | AnxB9:3R:21068755-21070916:1
5 | AnxB9:3R:21066678-21067194:1
6 | AnxB9:3R:21069279-21069753:1
7 | ash2:3R:24652748-24653376:1
8 | asp:3R:24760047-24760509:1
9 | aurA:3R:11964373-11964826:1
10 | ca:3R:29814401-29815047:1

[rj-patrick]

The output should look something like the following:

seqnames     start       end width strand gene_id UTR3 UTR5 intron exon CDS
Map2k7:8:4247528-4247926:1            8   4247528   4247926   399      +  Map2k7  YES                     
Arpc2:1:74262320-74268209:1           1  74262320  74268209  5890      +   Arpc2  YES                     
Tmem167:13:90098198-90105220:1       13  90098198  90105220  7023      + Tmem167  YES                     
Mpped2:2:106864471-106867623:1        2 106864471 106867623  3153      +  Mpped2  YES                     
Rab10:12:3249200-3253440:-1          12   3249200   3253440  4241      -   Rab10  YES                     
Psmc6:14:45342988-45349071:1         14  45342988  45349071  6084      +   Psmc6  YES                     
                               pA_motif pA_stretch pT_stretch Junctions
Map2k7:8:4247528-4247926:1        FALSE      FALSE      FALSE non-juncs
Arpc2:1:74262320-74268209:1        TRUE      FALSE      FALSE non-juncs
Tmem167:13:90098198-90105220:1    FALSE      FALSE      FALSE non-juncs
Mpped2:2:106864471-106867623:1    FALSE      FALSE      FALSE non-juncs
Rab10:12:3249200-3253440:-1       FALSE      FALSE      FALSE non-juncs
Psmc6:14:45342988-45349071:1       TRUE      FALSE      FALSE non-juncs

Can you confirm if you are reading in output from the AnnotatePeaksFromGTF function?

[aralbright]

Yep, there was a user error with the AnnotatePeaksFromGTF function, that fixed it. I'm having trouble now with DUTest.

> res_table = DUTest(peaks_seurat, 
+                    population.1 = colnames(peaks_seurat),
+                    exp.thresh = 0.1, 
+                    feature.type = c("UTR5", "UTR3"))
[1] "1564 expressed peaks in feature types UTR5, UTR3"
[1] "278 genes detected with multiple peak sites expressed"
[1] "758 individual peak sites to test"
Error in profile.set2[, i] <- sub.matrix : 
  number of items to replace is not a multiple of replacement length

I wouldn't be surprised if it were user error again considering this is the first time I'm working with Seurat objects, but I really do appreciate your help!

[rj-patrick]

Hi @aralbright,

DUTest requires either the name of cluster/cell-type or a vector containing a subset of cell IDs for the population.1 parameter. If you don't input anything for population.2 it will default to comparing to the remainder of the cells. However, It looks like you've selected all cells in your Seurat object by selecting colnames(peaks_seurat) - the result of this is that there are no cells to compare to, which is causing an error. Try using the name of a cluster instead and it should work.

Cheers,
Ralph

[aralbright]

Ahh okay, I wasn't sure what I should put there because I don't have assigned cluster names and in the end I'd like to look at all cells. I clearly need to do more work with Seurat objects so that I know what exactly I'm looking at, but I think I can take it from here. Thank you!

[rj-patrick]

Okay sure. Yeah Sierra is designed to look for relative differences between cell types, so try some clustering first then look for DTU differences with Sierra.

I will close this issue now, but feel free to reopen if you come across more problems.