Table 2 A brief overview of the steps required to make PRS relevant in a clinical setting

1. Realistic estimation of predictive ability in clinical populations, which may differ from research samples in disease severity, ancestral diversity, and exposure to environmental risk

2. Identification of the intended purpose of the PRS, which may affect its design and validation, and relevant clinical questions that can be answered, for example, prediction of severity, course of illness, or response to treatment

3. Recognition that even though not useful for the majority of the population with PRS in the middle of the distribution, the outcome may be relevant for those with high or low PRS, in the tails of the distribution

4. Clarification if PRS has an additive or interaction effect with established epidemiological or biological risk factors before combining in joint prediction models [88]

5. Engagement of clinicians and service users, to ensure that any application of polygenic risk scores avoids deterministic interpretations and is based on the understanding that PRS is an indicator, not a precise measure