Diffbind raw read counts

Question

[Beth]

Hi all,

I'm trying to obtain a raw read counts from Diffbind, using the following command:

H3K27ac<- dba.count(matrix_all, minOverlap=3,
                score = "DBA_SCORE_READS")

And then

peaks<-dba.peakset(H3K27ac, bRetrieve=TRUE, DataType=DBA_DATA_FRAME, writeFile = "H3K27ac.txt")

But the result is looking like as non-integer values, is it normal for this type of output (because, for, example, raw read counts is integer in RNA-seq data)?

Thank you!

Answer 1

[Rory Stark]

Yes, these should be integers when score=DBA_SCORE_READS.

I can not reproduce this in the currently released version (DiffBind_3.6.5) using the sample dataset:

tam <- dba(sampleSheet="tamoxifen.csv")
tam <- dba.count(tam, minOverlap=3, score=DBA_SCORE_READS)
peaks <- dba.peakset(tam, bRetrieve=TRUE, DataType=DBA_DATA_FRAME, writeFile="tam.txt")
peaks[1:5, 1:8]
read.table("tam.txt")[1:5,1:8]

Giving:

> peaks[1:5, 1:8]
    CHR  START    END BT4741 BT4742 MCF71 MCF72 MCF73
1 chr18 111395 111795     21     19   127    63   112
2 chr18 150222 150622     15     13    22    11    27
3 chr18 189192 189592     11     22    20    12    38
4 chr18 215629 216029     25     49    10    12    15
5 chr18 311744 312144     48     39    13    10     7
> read.table("tam.txt")[1:5,1:8]
     V1     V2     V3 V4 V5  V6 V7  V8
1 chr18 111395 111795 21 19 127 63 112
2 chr18 150222 150622 15 13  22 11  27
3 chr18 189192 189592 11 22  20 12  38
4 chr18 215629 216029 25 49  10 12  15
5 chr18 311744 312144 48 39  13 10   7

[Beth]

Thank you!