Missing some supporting reads

[helrick]

Hi there, many thanks for this great tool!

I have a use case where I'm trying to plot some long reads which support a deletion. In IGV, I can see 15 reads supporting this deletion. When I plot the same BAM using figeno, with the track settings of:

    {
      "type": "alignments",
      "file": <bam-path>,
      "label": "",
      "height": 50,
      "margin_above": 1.5,
      "bounding_box": true,
      "fontscale": 1.2,
      "label_rotate": true,
      "hgap_bp": 30,
      "vgap_frac": 0.3,
      "read_color": "#cccccc",
      "splitread_color": "#FF9F00",
      "link_color": "#999999",
      "link_lw": 0.5,
      "link_splitreads": true,
      "only_show_splitreads": false,
      "min_splitreads_breakpoints": 2,
      "group_by": "none",
      "color_by": "none"
    },

I am only seeing 5 reads:

In contrast to IGV (version 2.19.4), where I see 15:

As a result, the deletion appears to be of very low variant allele frequency, despite accounting for over half of the reads in this region.

One potential reason may be that some of the support appears to be from CIGAR string deletions rather than supplementary alignments. I've tried to plot some other deletions, which are only supported by CIGAR string deletions and have noticed that these deletions weren't shown in the reads - unless I'm missing a setting? Eager to hear your thoughts, though. Please let me know if I can provide more information.

All the best.

[e-sollier]

Hello! Indeed, figeno only considers as "split reads" reads that have several alignments in the BAM file, not reads that have a long deletion reported in the CIGAR string.

[helrick]

Thanks for clarifying! Are there any plans to add support for visualising long CIGAR string deletions and insertions in future?

[e-sollier]

No, sorry, I'm not planning to work on this.

[helrick]

No worries - I've opened a PR that attempts to implement this feature if you are interested in adding it