Integrative analysis of common and rare pathogenic variants for a more comprehensive genetic risk assessment

Abstract

Throughout this thesis, we examine the complex interplay between genetic factors, namely polygenic risk scores (PRS), rare pathogenic variants, and the impact of family history on the risk of complex diseases (e.g., breast and prostate cancer). Using their combined effect on cancer prevalence and lifetime incidence, we aim to demonstrate how they can be used to personalize cancer risk assessment. <br /> For the analyses conducted in this thesis, we have leveraged the large data of UK Biobank. At the time of these analyses, there were 200,643 samples available with both whole exome sequencing and genotyping data. This comprehensive dataset allowed us to classify individuals based on the carrier status of rare pathogenic variants in cancer susceptibility genes, if they have a high or low PRS (defined by 90th and 10th percentile thresholds), and whether they have a family history of cancer. Cox proportional hazards models were used to compute lifetime cumulative incidence of cancer, and multivariate logistic regression was used to compare odds ratios (ORs) across these groups. <br /> Based on genetic profiles of the individuals, the incidences of breast and prostate cancers have shown a distinct variation. For instance, compared to Individuals with lower PRS and absence of rare pathogenic variants, those with rare pathogenic variants and higher PRS exhibit a significantly higher cumulative incidence of cancer by age 70. Further, a family history of respective cancer increases the risk regardless of PRS. <br /> The findings of this thesis highlight the potential use of PRS in risk stratification approaches not only in the general population but also among individuals who carry rare pathogenic mutations. Breast and prostate cancer risks were shown to be influenced both independently and cumulatively by rare pathogenic variants, polygenic background, and family history. The thesis also highlights the urgent need for generalizability of PRS models across diverse ethnic backgrounds to enable more tailored and precise strategies for disease and cancer prevention.