What's the meaning of having different alignments with the same qname in a bam with no duplicates?

Question

I have removed duplicates from my bam file (not secondary, nor suplementary alignments), and then I have looked for some repeated qnames (to assess the failure of a program) and I've found a lot of them. All of them has TLEN equal to 0.

As an example: this is part of a bed file that a generate from the bam.

CHR  Start   End       QNAME                                 PNEXT  TLEN
12 86945793 86945944 A05325:240:SS348DSZX:2:5105:12246:3790 90044835 0
4 168461445 168461485 A05325:240:SS348DSZX:2:5105:12246:3790 90044835 0

And those are these reads in the original bam:

A05325:240:SS348DSZX:2:5105:12246:3790  2177    4       168461445       0       111H40M 11      90044835        0       ATAAAATTAAAAAAATAAAAAAATTAAAAAAAAAAAAAAA        :,F,F,,FF:FF,F,,F,FFFF,,F,F,:,:,FFFFFFFF     NM:i:2  MD:Z:28T8G2     MC:Z:151M       AS:i:32 XS:i:31 RG:Z:HM327DSXX.3        SA:Z:12,86945793,+,20S37M94S,1,1;
A05325:240:SS348DSZX:2:5105:12246:3790  129     12      86945793        1       20S37M94S       11      90044835        0       ATGGTGAAAAAACAAATTATTTAAAAAAATCAAAAAAAAAAAAAAAACCAAAAAACATAAAAAAAAGGGCAGTAATTTTATTCATGGGGTCAAACAAAAAAAAAGCATAATATAAAATTAAAAAAATAAAAAAATTAAAAAAAAAAAAAAA      FF,,,:,F,F,,,,F,,,,,,,F,FFFF:,:FFFF:FFF,FF::F,:,,FFFF,,,F,FF::,,F,::,,,,F,,F:FF,,,,,,,F::F,F,,,,F,FFFFF:,F,,F,,:,F,F,,FF:FF,F,,F,FFFF,,F,F,:,:,FFFFFFFF      NM:i:1  MD:Z:9G27       MC:Z:151M       AS:i:32 XS:i:29 RG:Z:HM327DSXX.3        SA:Z:4,168461445,+,111S40M,0,2;

This are the commands used to remove duplicates:

samtools view -F 1024 -bo filtered.bam original.bam

The commands used to generate this bed were:

samtools view sample.bam | awk '{ print $3, $4, $4+length($10), $1, $8, $9}' > sample.bed

Why do I present those alignments?

Answer 1

I think that you are dealing with two different phenomena:

1) TLEN == 0 reads

From the BAM spec:

TLEN: signed observed Template LENgth. If all segments are mapped to the same reference sequence, the absolute value of TLEN equals the distance between the mapped end of the template and the mapped start of the template, inclusively (i.e., end − start + 1).14 Note that mapped base is defined to be one that aligns to the reference as described by CIGAR, hence excludes soft-clipped bases. The TLEN field is positive for the leftmost segment of the template, negative for the rightmost, and the sign for any middle segment is undefined. If segments cover the same coordinates then the choice of which is leftmost and rightmost is arbitrary, but the two ends must still have differing signs. It is set as 0 for a single-segment template or when the information is unavailable (e.g., when the first or last segment of a multi-segment template is unmapped or when the two are mapped to different reference sequences).

Note that in each case the mate read appears to be mapped to a different chromosome/contig. So, the mapper can't estimate insert size in these cases, so it just says zero.

2. Multiple reads with same qname

There are various reasons why the same qname can appear multiple times. Again from the BAM spec, one such explanation is that it is a chimeric alignment:

Chimeric alignment: An alignment of a read that cannot be represented as a linear alignment. A chimeric alignment is represented as a set of linear alignments that do not have large overlaps. Typically, one of the linear alignments in a chimeric alignment is considered the “representative” alignment, and the others are called “supplementary” and are distinguished by the supplementary alignment flag. All the SAM records in a chimeric alignment have the same QNAME and the same values for 0x40 and 0x80 flags (see Section 1.4). The decision regarding which linear alignment is representative is arbitrary.

You may have to do some digging and looking at the different flags to decide what exactly is going on, but my bet is that they are supplementary alignments.