Question 3 of 9 Intermediate Concept

How do depth and breadth of coverage affect variant-detection confidence?

Short answer

Depth is the number of reads covering a position, while breadth is the fraction of target bases covered sufficiently well. Both matter. A sample can have excellent average depth while important regions have too few informative reads for reliable detection.

Why

Depth describes how many reads overlap a site, and breadth describes how much of the intended region reaches a defined coverage threshold. Confidence is not a function of raw depth alone. Reads must also have sufficient base and mapping quality, represent independent DNA molecules and support the expected allele fraction.

For example, a panel might average 100-fold coverage but leave an exon at only three-fold due to GC bias or capture failure. Variants in that exon may be missed even though the run average looks strong. At a covered heterozygous germline site, allele balance and strand representation help assess confidence.

For low-frequency somatic variants, error rates and molecule counts become more important and the required coverage is generally assay specific. Report thresholded breadth, depth distributions, uniformity, low-coverage intervals and validated detection limits. A negative result in an insufficiently covered region means not assessable, not definitely wild type.

Common mistakes

Using mean depth as proof that every clinically relevant target base was confidently covered, or counting duplicate reads as independent support

The interviewer might ask next

What would you do if a high-depth run contains a poorly covered disease-relevant exon?