HomeCLC software: Important notificationsIssues affecting only older versions of productsVariant calls possible within common sequence regions of QIAseq DNA panel results under specific circumstances

2.22. Variant calls possible within common sequence regions of QIAseq DNA panel results under specific circumstances

Issue description

For paired QIAseq panel data, common sequence is left on the end of the non-indexed read when that read is longer than the DNA fragment being sequenced.

This will usually not cause problems, as the contaminating common sequence is not aligned to the reference by the read mapper, and so is ignored by variant detection tools. However, false positive variants may on occasion be called when i) enough of the common sequence, by chance, matches the reference, leading to alignment and ii) there is at least one mismatch in that region, leading to a variant call.

We expect false positive variants in such regions to occur rarely in practice due to the combination of circumstances needed. See the Further Background section below for more information.

This issue is caused by a problem with the Remove Ligation Artifacts tool. This tool, and workflows that include this tool, are distributed in the QIAseq DNA V3 Panel Analysis and the QIAseq Targeted Panel Analysis plugins. A full list of software and versions affected is provided at the end of this article.

 

Further background

When the length of a sequencing read exceeds the length of the DNA fragment being sequenced, the read can end with artifacts from library preparation, such as adapter sequence. 

In the case of paired QIAseq V3 panel data, an artifact routinely encountered is "common sequence" at the 3' end of the non-indexed read.  This is shown in simplified form below.  If the read is long enough, it may also contain some of the UMI sequence.

The Remove Ligation Artifacts tool trims away artifacts, so that they are not considered in variant detection analysis. However, in the affected software, regions of common sequence and UMI sequence remain on the non-indexed read of a pair. This is generally not a problem, as the common sequence and UMI do not usually match the reference where the read has mapped. Non-matching regions in read mappings are ignored when variant detection analysis is run.

However, if the part of the common sequence adjacent to the DNA fragment, by chance, matches the reference well enough to be mapped, variant detection tools will consider that region. At least 2 bases of the common sequence near the target region must match the reference for this to happen with default read mapping penalties. If such a region matches perfectly, then this does not affect the variant detection results as no variants will be reported. However, if at least one base does not match the reference, then a variant can potentially be called within the common sequence region. 

Examples

A false positive variant, in the common sequence at the 5' end of the DNA fragment of interest:

The CCT within the red box is part of the common sequence, but the 2 Cs happen to match the reference at this point. This leads to these bases being mapped to the reference and being considered during variant detection analysis.  The T of the CCT does not match the reference, and here has been called as a variant by the variant detection tool.

 

A false positive variant, in the common sequence at the 3' end of the DNA fragment of interest:

The AGGA in the red box is part of the common sequence.Three of the four bases in this region happen to match the reference, so this region is mapped to the reference and is thus considered during variant detection analysis. The first G in this region does not match the reference, and in this case is called as a variant by the variant detection tool.

 

Other symptoms

When considering if a particular variant is a false positive due to this issue, it can be useful to consider that the common sequence will appear on only one of the pair members - either only the ones that map in the reverse direction to the reference, OR only the ones that map in the forward direction to the reference. To check for this, choose the option in the Reads Track section of the side panel called "Disconnect paired reads" and check if all the affected reads are red or all of them are green.

 

Affected software

The Remove Ligation Artifacts tool and the QIAseq DNA v3 Panel Analysis workflow distributed in:

  • QIAseq DNA V3 Panel Analysis 1.0  (plugin)
  • QIAseq DNA V3 Panel Analysis Server Plugin 1.0

The Remove Ligation Artifacts tool and the Targeted DNA and Targeted RNAScan workflows distributed in:

  • QIAseq Targeted Panel Analysis 1.0  (plugin)
  • QIAseq Targeted Panel Analysis Server Plugin 1.0

The Detect QIAseq RNAscan Fusions (beta) workflow distributed in:

  • QIAseq Targeted RNAScan Panel Analysis 0.5.1 beta 1
  • QIAseq Targeted RNAScan Panel Analysis  Server Plugin 0.5.1 beta 1

 

This issue was fixed in

  • QIAseq Targeted Panel Analysis 1.0.1
  • QIAseq Targeted Panel Analysis Server Plugin 1.0.1
  • QIAseq DNA V3 Panel Analysis 1.0.1
  • QIAseq DNA V3 Panel Analysis Server Plugin 1.0.1
  • QIAseq Targeted RNAScan Panel Analysis 0.5.2 beta 1
  • QIAseq Targeted RNAScan Panel Analysis Server Plugin 0.5.2 beta 1

This page was: Helpful | Not Helpful