There comes a point during the development of every NGS assay at which you want to make sure it will reliably detect everything you say it can. You want its real-world performance to match your claims.
That means fine-tuning your development protocol — from specimen handling, to nucleic acid extraction, to library prep — against a range of variants and allele frequencies to make sure your early-stage assay picks them all up.
Only after doing so can you move the assay on to the costly validation phase.
Naturally, to replicate real-world scenarios, many developers instinctually turn to real patient samples.
But the truth is, if you’re only subjecting your assay to the patient samples your lab has on-hand — or even ones you procure from colleagues or biobanks — you’re not exposing your assay to a wide enough range of variants and conditions to ensure its performance. Plus, you may be costing your lab money and time:
- Development delays can allow another lab to launch their assay before yours.
- Not using robust enough reference materials can lead to your assay not meeting performance claims in production.
While there is clearly a need to evaluate assays with real specimens tested on orthogonal methods, in the development phase multiplexed truth sets such as biosynthetic NGS reference materials are superior to patient specimens.
Here are three reasons not to trust remnant patient samples alone for assessing your clinical NGS assay’s workflow.