4.6 years ago by
Walnut Creek, USA
I don't know of a tool that does this, specifically. However, you could accomplish it in a slightly different way:
Adapter-trim all of the reads.
Merge all of the overlapping reads, generating a file of merged reads, and another file (or pair of files) of the unmerged reads.
Fuse together all of the remaining unmerged pairs: this means reverse-complement read 2, and then concatenate the sequences. e.g. if read 1 is AAAA and read 2 is CCAG, the fused read would be AAAACTGG. I don't have a tool for that, sorry!
Concatenated the merged reads with the fused reads, then map all of them with BBMap, which can handle the resulting potentially long indels at the junction where the non-overlapping pairs were fused together. Using the "covhist" flag, it will directly output the frequency of base coverage. This will consider bases under by deletion events to be covered, so it will calculate the physical coverage including the unsequenced portion:
bbmap.sh in=reads.fq ref=reference.fa nodisk covhist=covhist.txt