Determine relative position (start and end) of chimeric secondary alignment within original long read

Question

[mattia.bosio85]

Hi all,

I am relatively new to this field and I have not been able to find an answer yet. I have long reads from nanopore and aligned with BWA mem -x ont2d and I see several are chimeric reads with reported secondary alignments (SA flag in the BAM file)

I would like to determine where the chimeric alignment lies within the original read. I guess it could be obtained from the CIGAR code but it looks pretty complex at a first glance. Is there an algorithm to follow to recover it ? My main goal is to see if I can reconstruct where each piece of the original read lies in the reference genome

Here an example of a SA item :

SA:Z:chr9,10158,+,32S6M1D6M1D6M1I12M2D3M1D4M2I7M1D15M2I7M1D4M2D7M1D16M1I1M1I2M 2I10M3D7M2I9M1I11M4I5M2D3M1D15M2D11M2I1M2I18M1I22M1I7M3I16M2D4M1D23M3D12M1D3M1I 5M1D7M3D30M1I8M2D23M1D39M2D23M1D13M3D2M1D5M2D5M1D14M2D29M2I15M1D22M1D 1M3D26M2D15M2D3M3D21M1D7M1D18M1D3M1D6M2I22M1D14M1D3M1D20M1I6M1D12M2D7M 2D39M1D1M1D9M2D8M1I9M1D1M1D2M1D6M1D16M1D9M1I14M1D4M1D4M3D4M1D3M1D20M1I24M 1I1M3I3M3I7M1I22M1D3M1D16M22277S,44,260;

• I understand that here the first 32 bases and the last 22277 are soft clipped. Is it correct to say that the secondary alignment starts at position 32 and ends at length(read) - 22277 ? How should I deal with the 'D' or 'I' parts within the CIGAR value ?

Thanks a lot for your input about this

Mattia

[Vitis]

This is very interesting. I dealt with CIGAR string a lot but have never seen this complexity. Can you check the primary alignment? Did the clipped 22277bp mapped to somewhere else?

[WouterDeCoster]

How was the library generated? If (long range) PCR was used you can get chimer molecules as an artefact of the library prep. Although it sounds rather unlikely that you have a PCR product of > 22277bp.