Examining the relationship between year of birth and polygenic risk scores

Polygenic risk scores (PRS) are already being used in the clinical context (e.g., cancer genetics) and they may eventually be implemented in routine clinical practice as clinical biomarkers.

It is both important and urgent to develop a better understanding of how non-genetic factors biology may affect the value of the PRS, before applying them in clinical settings.

In BioVU, we observed relationships between year of birth and PRS trained on a number of phenotypes including Attention Deficit Hyperactivity Disorder, Major Depressive Disorder, height, and educational attainment. These relationships were also replicated in MGB and MVP.

This proposal is inviting a working group to identify and evaluate factors influencing the relationship between PRS and year of birth in biobank cohorts.

We wish to accomplish the following aims:

1. Test whether the correlation between year of birth and PRS value is replicated in other PsychEMERGE sites,
2. evaluate the impact of changing population substructure and
3. ascertainment biases into the medical system as possible explanations for the observed correlation.