picard – scatter intervals by n uses

Question

[Matteo Ungaro]

Hi there,

I have a FASTA genome (GRCh38) for which I want to detect and output a BED file containing intervals of the N sequences in the same. It appears Picard has a functionality to do so — scatter intervals by Ns; however, I'm unsure whether this is actually doing what I need.

In practice, the command below results in a one-based file, as opposed to the standard zero-based format of BEDs...; therefore, if someone has more experience, I would like to know whether and how I can use this output file with bedtools to selectively subtract these regions/intervals from the BED coordinate for the entire genome.

Thanks in advance!

java -jar picard.jar ScatterIntervalsByNs \
      R=hg38.fna \
      OT=N \
      O=hg38_one.intervals

Answer 1

[Pierre Lindenbaum]

https://gatk.broadinstitute.org/hc/en-us/articles/360036453012-IntervalListToBed-Picard

Trivially simple command line program to convert an IntervalList file to a BED file.

[Matteo Ungaro]

@Pierre Lindenbaum, I see. Essentially, what it does is removing the header and subtracting 1 from the first column? Just to make sure because my more straightforward approach would have been to use grep -v and awk trying to accomplish the same. Let me know, thanks!

[Pierre Lindenbaum]

Essentially, what it does is removing the header and subtracting 1 from the first column

yep :-)