I’m not entirely sure what you are actually trying to do. Are you trying to add this peakset to use as a consensus peakset for counting? If so you can specify it as a parameter to
dba.count() by setting
peaks=merged_narrowPeak_sorted without having to load it to look like a sample (which it’s not). Alternaitvely, if you have a bunch of these files, all with the same merged peakset but different counts pre-computed, you can add them via
dba.peakset() using the
counts= parameter, with the count values in the fourth columns (e.g.
If you really do need to create a new
DBA object with a consensus peakset as a sample, there are some with
dba.peakset(), which I’ll sort out soon. In the meantime I can give you some ways to get past this.
First, you should be setting
peak.caller only sets a metadata string, nothing more.
However there is a bug right now in doing this. There are a number of alternative workarounds you can try while a bugfix is pending:
- You can pass the bedfile instead of the table:
- You can set
- You can pass the score in column 4:
In all of these cases, there is a bug when adding an initial peakset using
DBA=NULL. You have to add a line after the first call to
dba.peakset(). for example:
consensus.peaks <- dba.peakset(NULL,peaks="merged_narrowPeak_sorted.txt"`, peak.format = "bed") consensus.peaks <- dba(consensus_peaks)
I’ll be addressing the bugs soon in maintenance release.
Read more here: Source link