Quality Tab

The Quality section gives you a clear, interactive overview of variant quality across all sequenced individuals in a case. This includes zygosity, quality grades, and key per-variant metrics—helping you assess data confidence and review evidence effectively.

Visual Family Tree with Zygosity and Quality

• Each sample in the pedigree is represented with:

  • Zygosity status (HET, HOM, HEMI, REF) shown inside the pedigree symbol.

  • Variant quality grade—High (H), Medium (M), or Low (L)—shown next to the symbol.

• The proband’s quality grade is also highlighted in the section title, making it easy to spot key issues.

• The title includes the variant caller notation, indicating the source of the variant call.

• Click on a family member’s icon to switch the sample in view.

Quality Metric per Variant Type

Each variant is evaluated based on a tailored set of quality metrics, which are displayed for every individual who carries the variant.

1. Zygosity

Determines whether the variant is present on one allele (heterozygous), both alleles (homozygous), or the X chromosome in males (hemizygous) or absent. This is critical for inheritance pattern checks.

DRAGEN genotyping calculates this status per sample using base calls and alignment data with confidence scores (genotype quality, coverage, etc.), which supports downstream variant quality grading.

2. Variant quality grade

Overall confidence category (High, Medium, Low) based on caller-specific rules for QUAL score, FILTER status, and other QC tags.

QUAL is a Phred-scaled score representing variant-caller confidence. High quality typically reflects robust support from sufficient read depth, alignment agreement, and noise filtering.

3. SNV/Indel metrics:

• Base Quality (BQ) — Average Phred score for bases supporting the variant. High BQ = more reliable base calling.

• Depth (DP) — Total number of reads covering the position. It is the count of high-quality (MQ>0) aligned reads post-duplicate/secondary removal.

• Mapping Quality (MQ) — Confidence in read placement on the reference genome (Phred-scaled).

• Genotype Quality (GQ) — Confidence in zygosity call. It is derived from likelihood ratios between genotypes, scaled as Phred.

4. CNVs metrics:

• Copy Number — Estimated number of copies in the region.

• CNV Quality Score — Confidence score from caller.

• Size — CNV length (bp).

• Bin Count — Number of read-depth bins spanning the CNV.

• Likelihood Ratio Score (LRS) (DRAGEN 4.3, exome/panel CNVs only) — log₁₀ ratio of ALT vs REF probability.

5. STRs metrics:

• Depth — Coverage at repeat locus.

• Repeat Number — Count of repeat units.

• Repeat Length — Base-pair length of repeat tract.

• Confidence Intervals — For REF and ALT alleles, showing repeat size accuracy (DRAGEN 4.3+).

Special case: STR loci ARX & HOXA13 are always marked Low quality.

6. Structural variants (SVs):

SV callers use breakpoint clusters, depth shifts, and split-read evidence to compute QUAL. Large SVs sometimes fail quality thresholds despite real biology being responsible.

7. MRJD caller variants:

Highlights calls in difficult-to-map regions using parologous read mapping logic. MRJD aggregates haplotype evidence across duplicated regions; JIDS tags group ambiguous mappings; REGION_AMBIGUOUS flags uncertain assignments

8. Targeted caller variants:

For targeted genes with complex mapping (e.g., CYP21A2, GBA, HBA). Targeted caller uses specialized algorithms handling high-similarity regions; JIDS then references ambiguous mapping; recombinant tags reflect complex locus structures.\

9. Allele fraction and Distribution:

• For SNVs/Indels, allele fractions are shown as pie charts—one per sample—offering an intuitive snapshot of variant distribution across alleles. ALT and REF counts pulled from genotype likelihoods for used for display.

• For STRs, the allele distribution table shows how sequencing reads support an STR variant, helping you evaluate call accuracy. The columns are denoted by REF and ALT – represent reference and alternate allele counts. Rows correspond to three categories:

  • In Repeat – Reads aligning within the repeat (higher values improve confidence)

  • Flanking – Reads adjacent to the repeat region.

  • Spanning – Reads covering the entire repeat region.

• This feature is not applicable to CNVs.

To change the sample in review for the particular variant, click on the corresponding icon in the family tree.

Use the quality tab

• Before interpretation — to verify if a variant meets minimum quality thresholds.

• For cross-sample comparisons — to check if the variant is consistently high-quality across proband and relatives.

• To filter analysis — by excluding low-quality calls that may be sequencing artefacts.