Suppose I have two groups: treatment & control. Each group has a value assigned to each interval on the genome. Let me explain with the following example:
File A contains data for the treatment group and has the following coordinate sorted format:
chr1 100 110 3
chr1 200 212 8
chr1 220 240 5
chr1 500 600 1000
...
File B contains data for the control group:
chr1 100 108 7
chr1 150 160 14
chr1 210 230 88
chr1 700 850 3
...
And I would like to get results like the following: chr1 100 110 3 7
chr1 150 160 0 14
chr1 200 240 13 88
chr1 500 600 1000 0
chr1 700 850 0 3
...
The 3rd entry is a merge of two tracks on A sided by an entry on B.
Is there a software product to do so?
Thank you!
•
written
by
moxu
3 answers
I think this may be able to do it: standardize the windows on the left side, instead of using jumping windows, pick a window size and lay them out across the entire genome (such as chr1 1 10, chr1 11 20, chr1 21, 30...), leaving some windows with zero values. Then it might be easier to merge your data as you like.
Well, the solution based on bedops below is perfectly fine for this type of problem, I usually create this type of data with seamless chromosome windows for data standardization and documentation purpose. Somewhere down the road, you may repetitively need these operations to intersect new data. The standardized windows would make your life much easier in the future.
Via BEDOPS:
Merge sets A and B with bedops:
$ bedops --merge A B > AB.merge
Map set A to the merged set and sum the ID values with bedmap and awk:
$ bedmap --echo-map-id --multidelim '\t' AB.merge A | awk '{if(NF==0){print 0;}else{s=0;for(i=1;i<=NF;i++){s+=$i;}print s;}}' > A.merge.sum
Repeat with set B:
$ bedmap --echo-map-id --multidelim '\t' AB.merge B | awk '{if(NF==0){print 0;}else{s=0;for(i=1;i<=NF;i++){s+=$i;}print s;}}' > B.merge.sum
Paste the AB merged set to the A and B merge-sums with paste:
$ paste AB.merge A.merge.sum B.merge.sum
chr1 100 110 3 7
chr1 150 160 0 14
chr1 200 240 13 88
chr1 500 600 1000 0
chr1 700 850 0 3
Log in to answer this question.
-
bedtools annotate
written by alexmondaini •
-
Plot assigned / unassigned reads distance to 3'UTR
written by nlehmann
-
Finding the start codon
written by Mimmi Ahlmén •
-
bedtools what in B absent from A
written by Medhat
-
Merge bed files, keep the peak count
written by rioualen
-
needleman-wunsch (ggsearch implementation) not symmetric
written by comacke •
-
Bedtools intersect A contains B and vise versa
written by Medhat
-
Count Columns For Each Window
written by roll
-
Which Of The Genes Are Enriched With Repeat Elements
written by roll
-
How To Separate A List Of Genomics Regions Into Several Disjoint Subsets
written by dustar1986
