Discrepancies between pysam and wally

[MJoseMo]

Hi, I've been using pysam to extract positions in my target sequence. However, I've encountered some discrepancies when visualizing these alignment positions with Wally. Some positions are not equal. Did I need to consider the orientation of alignments when working with pysam or did it encounter problems when extracting positions of long-read sequences

[tobiasrausch]

For wally, forward matches are in blue, reverse matches in orange. I haven't used pysam.

[MJoseMo]

Hi, my concern lies in the fact that both procedures should yield similar results

[tobiasrausch]

Does pysam report relative to the match (just the given SAM record) or relative to the primary alignment?

[tobiasrausch]

Can you try this with pysam:

echo -e ">ref\nATCTTTGGTGTCTGTGTGTGTGCGCGCGCACGTGCATGTGTGTGAGTTGTCATTGTTTTTGTTTGTTTTG\nTATTCCTCAGAAATAGAAAGCTCCCATTAGG" > ref.fa
echo -e ">query\nATCTTTGGTGTCTGTGTGTGTGCGCGCGCACGTGCATGTGTGTGAGTTGTCATTGTTTTTGTTTGTTTTG\nCCTAATGGGAGCTTTCTATTTCTGAGGAATA" > query.fa
bwa index ref.fa
bwa mem ref.fa query.fa | samtools sort -o query.bam -
samtools index query.bam
wally matches -x 600 -r query -g ref.fa query.bam

[MJoseMo]

Dear @tobiasrausch, I get the following results:

Forward:

Reference Name: ref Reference Start: 0 Reference End: 70 Query Start: 0 Query End: 70

Reverse:

Reference Name: ref Reference Start: 70 Reference End: 101 Query Start: 0 Query End: 31

[MJoseMo]

Dear Tobias, I know it's been a while since my last interaction, but I still don't understand the discrepancies between wally's and pysam's results, specially for reverse matches.