From a variant to a drug claim
Evaluate a molecular drug claim from credible finding through measured biology to clinical evidence and access.
A sequencing report can offer a useful clue. The next step is to find out what the clue actually supports. This guide helps you ask a sequence of concrete questions without needing to become a specialist in every pathway.
The core path takes about 46 minutes. Start with DNA, RNA, and protein if those layers are unfamiliar. The multi-omics guide offers a fuller introduction to connecting measurements.
The big ideas
- A finding needs a clear specimen, assay, and attribution before it can support a biological claim.
- A gene name does not establish the direction of a protein or pathway change.
- A measured state and a selective survival dependence are different conclusions.
- Clinical evidence must fit the disease, biomarker, treatment setting, and intended outcome.
- Eligibility and access are additional gates; neither substitutes for evidence of benefit.
The map
Lessons
| Step | Lesson | The question it answers | Minutes | Learning aid |
|---|---|---|---|---|
| 1 | Authenticate and attribute the finding | Is the finding credible, and which cells carry it? | 14 | A read-fraction and specimen exercise |
| 2 | Connect repair biology to PARP sensitivity | What consequence was measured, and what remains predicted? | 16 | Two models with similar scars and different current repair |
| 3 | Check clinical evidence and access | Does the evidence fit the intended use? | 16 | Disease, biomarker, and setting comparisons |
The repair example teaches a general method. It does not make every drug claim a homologous recombination deficiency (HRD) question. Use the pathway mechanisms library when your proposed consequence involves recycling, growth signals, lipid supply, developmental signals, or stress defenses.
Concepts you will use
For the finding, use mutation types, germline versus somatic, and variant allele fraction versus cancer-cell fraction.
For the consequence, use homologous recombination and HRD and synthetic lethality. The supporting library links to its specialist concepts at the start of each lesson.
For translation, use clinical actionability, prognostic versus predictive biomarkers, and analytical validity, clinical validity, and utility. A clear mechanism and a valid measurement still leave a clinical comparison to answer.
Keep an evidence ledger
At each step, record the observation, its source, the conclusion it supports, and the next unresolved question. A fictional repair-gene variant may be confirmed in DNA, alter an RNA junction, and accompany a protein signal. A functional assay then asks whether repair is impaired. A clinical study asks a different question about benefit in a defined population.
From data to a care plan is the shared supporting lesson for this ledger. Judging a test or vendor claim helps assess the claim's wording and validation. Accessing experimental therapy develops the practical final gate.
Applied to Diana
The molecular profile owns qualified findings. The postoperative treatment question owns the care decision. Use these owners for application; this guide teaches the reasoning rather than maintaining case results.
What is still uncertain
A model can differ from the tumor in cell mixture, exposure, and environment. A pathway can be active without being a selective bottleneck. A clinical signal can be real while leaving an unstudied setting unresolved. Each uncertainty should stay attached to the claim it limits.
Source check: 2026-10-09. General education; expert and learner review pending. Study and indication examples are dated in the lessons and need a fresh check for a real clinical decision.