Problem with cigar in getNiceAlignment

[francoisfauteux]

Hello, the expression under "check CIGAR correct" in getNiceAlignment "^[XDI1-9=]*$" does not account for all possible cigar ops according to https://samtools.github.io/hts-specs/SAMv1.pdf (1.4.6. CIGAR: CIGAR string) it could be "^[MIDNSHPX0-9=]+$". First test below works, second example does not. Maybe this can also help: https://rdrr.io/bioc/GenomicAlignments/man/cigar-utils.html. Thank you.

library(edlibR)

qry<-"this is a test"
trg<-"test this is a"
res=align(qry,trg,task="path")
res$cigar
#[1] "1=5D8=5I"
getNiceAlignment(res,qry,trg)
#$query_aligned
#[1] "t-----his is a test"
#$matched_aligned
#[1] "|-----||||||||-----"
#$target_aligned
#[1] "test this is a-----"

qry=paste(rep("this is a test",100),collapse=" ")
trg=paste(rep("test this is a",100),collapse=" ")
res=align(qry,trg,task="path")
res$cigar
#[1] "1=5D1493=5I"
aln=getNiceAlignment(res,qry,trg)
#Error in getNiceAlignment(res, qry, trg) :
#The CIGAR string is in an invalid format. Operation detected is not a single character '=' or 'X' or 'D'.
#The expected format should be 'number of occurrences' + 'CIGAR operation', e.g. '4=' or '1D5=1X1=1X'.Please fix.

[evanbiederstedt]

Hi @francoisfauteux

Thanks for the issue.

I was simply was using the same CIGAR parsing operations within the python library: https://github.com/Martinsos/edlib

Please create a pull request to incorporate other CIGAR parsing operations if you wish; it could be incorporated in both libraries.

@Martinsos may have more thoughts here.

Best, Evan