Entering edit mode
hkitano
•
0
@hkitano-23077
Last seen 4.1 years ago
I have 10 "summit" bed files from macs2 peak-calling, with 5 factors with 2 replicates each.
I added them all one-by-one to a dba object "dbObj" that looks like this:
10 Samples, 3202 sites in matrix (234027 total):
ID Factor Replicate Caller Intervals
1 CTR-1 ctr 1 bed 29940
2 CTR-2 ctr 2 bed 27812
3 IFNG-1 ifng 1 bed 20954
4 IFNG-2 ifng 2 bed 19747
5 IL4-1 il4 1 bed 25353
6 IL4-2 il4 2 bed 25543
7 preIL4-1 preil4 1 bed 21657
8 preIL4-2 preil4 2 bed 20200
9 preIL4posIFNG-1 preil4posifng 1 bed 24002
10 preIL4posIFNG-2 preil4posifng 2 bed 22068
Then, I tried to follow the steps in the answer here to deal with the replicants. I called:
dbObj2 <- dba.peakset(dbObj, consensus = -DBA_REPLICATE)
dbObj2 <- dba(dbObj2, mask=dbObj2$masks$Consensus)
and got this for dbObj2:
5 Samples, 0 sites in matrix (416 total):
ID Factor Replicate Caller Intervals
1 ctr ctr 1-2 bed 112
2 ifng ifng 1-2 bed 76
3 il4 il4 1-2 bed 77
4 preil4 preil4 1-2 bed 63
5 preil4posifng preil4posifng 1-2 bed 88
which looks right to me. But if I try this:
ConsensusPeaks <- dba.peakset(dbObj2, bRetrieve=TRUE)
ConsensusPeaks is a null object. I'm trying to be able to use dba.counts() on the 5 factors, is there any other way or am I doing this wrong? Thanks!
Thank you for replying, Rory. When I try to read in a
*_peaks.xlsfile, I get the following error:Any ideas for what this means? Thank you!
Hugo
You need to use:
Thank you! I appreciate your help. Going back to this ticket, I'm having trouble reproducing your results, and keep running into the same
Can't count: some peaksets are not associated with a .bam file.error.I have my dbObj, which has 10 peaksets (2 replicates for 5 factors)
The problem is when I run the following, I get the can't count error on the last line of code
I feel like I'm doing everything you suggested...do you see an issue? Thank you!
Hugo
If you send me the
dbObj, I can look and see what is going on.-Rory
Here's how I build my dbObj:
I do this 5 times for 5 factors, and end up with the following dbObj:
Then, I run the following code, and fail on the last line (
Error in pv.counts(DBA, peaks = peaks, minOverlap = minOverlap, defaultScore = score, : Can't count: some peaksets are not associated with a .bam file.).If I do the following to check for bam files, i get NULLs
I think I'm confused about the process of building DBA objects via peaksets. I have bam files associated with each .xls file from before I called peaks on each replicate. Are including those necessary when I instantiate each peakset for the DBA object?
Yes, you must tell
DiffBindwhere the bam files are for each peakset so it can count reads when you invokedba.count(). You can do this using thebamReadsparameter todba.peakset(), but in general it is much easier to keep track of this using a sample sheet passed to a single call todba().