How analyse chip-exo paired read data and get BigWig or Peaks called?

Question

[morovatunc]

Dear all hi,

I have paired end chip-exo data. The interesting is, both reads from each of the pair was already mapped and peaks were called. I have data in BED format which caused a confusion because of the paired end reads. Even though I have found couple softwares for paired chip-exo analysis, they all require bam input. I dont want to do all the alignment again therefore I wanted ask the community of booster.

My aim is to get BigWig file format for visualisation and ~250 bp peaks for the data process(finding overlaps etc.);

1) I am totally lost for the bigwig part so I beg for your information. Can I obtain BigWig with current available information? (I checked bed to BigWig threads from seq answers and biostars but can I solve it with using bedtools genomeconv)

2) For obtaining ~250 bp peaks I thought merging them with bedtools could be a solution. Is there a gold standard for merging process ?

Could you enlighten me a bit. Any document regarding this information would be very useful as well.

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE43785 — > source of the data.

            Tuncs-MacBook-Pro:~ morova$ more /Volumes/blackhole/chip-exo/GSE43785_RAW/GSM1071284_120424_ChipEXO5_NoIndex_L005_R1.fastq.tagAlign.bed
            chr12   79171842        79171891        -
            chr4    144341449       144341498       -
            chr3    42101448        42101497        -
            chr19   18496785        18496834        +
            chr3    24961259        24961308        +
            chr5    4818754 4818803 +
            chr10   88177242        88177291        +
            chr2    45031538        45031587        +
            chr3    190478599       190478648       +
            chr20   8807491 8807540 -

Thank you for your help,

Best,

Tunc.

[dyollluap]

ChIP-exo data probably won't have 250bp peaks, resolution is much better than ChIP-seq - see fig from Rhee & Pugh: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3243364/figure/F1/

Answer 1

[Ian]

I think you need to remap to get the BAM files (for genomecov). The description for each sample implies that the bowtie mapped reads are included, presumably in GSE43785_RAW.tar (~60Gb). If you have the space it would be worth checking.

[morovatunc]

Thank you very much for your reply!

Answer 2

[genecats.ucsc]

Hi Tunc, you can indeed convert bed files to bigWig files using a few UCSC utilities. The main ones you will be interested in are the bedItemOverlapCount and bedGraphToBigWig utils. You can find the appropriate pre-compiled binaries for your system here: http://hgdownload.soe.ucsc.edu/admin/exe/

The usage message from bedItemOverlapCount then explains what to do:

$ bedItemOverlapCount             
bedItemOverlapCount - count number of times a base is overlapped by the
    items in a bed file.  Output is bedGraph 4 to stdout.
usage:
 sort bedFile.bed | bedItemOverlapCount [options] <database> stdin
To create a bigWig file from this data to use in a custom track:
 sort -k1,1 bedFile.bed | bedItemOverlapCount [options] <database> stdin \
         > bedFile.bedGraph
 bedGraphToBigWig bedFile.bedGraph chrom.sizes bedFile.bw
   where the chrom.sizes is obtained with the script: fetchChromSizes
   See also:
 http://genome-test.cse.ucsc.edu/~kent/src/unzipped/utils/userApps/fetchChromSizes

If you have any follow up questions, it would be helpful if you could post them to our Google Groups forum: https://groups.google.com/a/soe.ucsc.edu/forum/#!forum/genome, that way our whole team can see the question and help with an answer.

Thanks,

ChrisL from UCSC

[morovatunc]

Thank you very much for your reply!