Graded recommendations, not raw genotypes

A list of variants is data, not advice. A prescriber at the point of care does not need another table of diplotypes to decode — they need to know which drug to avoid, which dose to adjust, and how strong the evidence behind that call is. A report that stops at the genotype hands the hardest interpretation to the busiest person in the room.

A defensible report grades every recommendation against published guidelines. SignalPGx evaluates each gene-drug pair against CPIC guidelines, FDA pharmacogenomic biomarker labeling, DailyMed prescribing data, and DPWG, so every actionable statement carries an explicit strength of evidence instead of an implied one.

A medical director signs every report

Automated interpretation on its own is not a clinical opinion. Software can surface a gene-drug interaction reliably, but a report only becomes defensible when a qualified clinician takes responsibility for what it says.

SignalAI assists interpretation and flags gene-drug interactions, yet the lab's own medical director reviews and signs out every report. That sign-out makes each result:

• signed

• accountable

• auditable

This keeps a licensed human in the loop and gives the ordering physician a report they can stand behind.

Transparent, traceable sourcing

A recommendation is only as trustworthy as the evidence behind it. When a report cannot show where a call came from, a clinician has no way to weigh it against their own judgment.

SignalPGx reconciles more than 15 curated sources into every interpretation and cites the guideline behind each recommendation. Coverage spans 50 pharmacogenes and roughly 950 medications, so a reviewer can trace any statement back to CPIC, FDA labeling, DailyMed, DPWG, or PharmGKB rather than take it on faith.

Staying current as the science moves

Pharmacogenomic evidence does not stand still. Guidelines are revised, new biomarkers earn labeling, and a patient's medication list changes over time.

A report signed last year shouldn’t quietly become stale. Living Reports automatically reanalyze a result when guidelines or a patient's medications change, and any report the sweep flags is routed back for a fresh medical-director review.

From result to defensible report

The lab runs the genotyping; SignalPGx turns the output into a report the EHR can consume.

One pipeline takes VCF, CSV, Agena MassARRAY, or array output and normalizes it into a report.

Genotypes are called, matched to guidelines, and assembled into branded, physician-reviewed output, then delivered over FHIR R4, HL7 ORU, CDS Hooks, or SMART on FHIR. A revocable Patient Passport QR link lets a patient share a minimal-PHI view the lab can revoke.

decision = model.predict(input_data)
execute(decision)

What defensibility looks like in practice

A defensible report is legible at a glance. A prescriber should be able to open it and see not just the finding, but the reasoning and the review behind it.

A strong report makes clear:

1. which drugs are affected

2. how strong the evidence is

3. who reviewed and signed it

Instead of guessing, a clinician can act on the report with the ordering decision firmly in their hands.

Conclusion

A defensible PGx report doesn’t just report a genotype — it stands behind the recommendation.

Labs that treat the report as a clinical document, not a data export, deliver something a prescriber can trust: graded, sourced, physician-reviewed guidance that stays current. SignalPGx is not a diagnostic test and is not FDA-cleared, and final prescribing always stays with the ordering clinician.