Disease risk prediction with rare and common variants

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Previous research indicates common genetic variants offer limited predictive value for complex diseases beyond established risk factors.
• Advancements in sequencing technology enable the identification of rare genetic variants with potentially greater functional impact.

Purpose of the Study:

• To evaluate the predictive value of rare genetic variants in complex disease risk assessment.
• To compare the performance of prediction models with and without the inclusion of rare variants.

Main Methods:

• Utilized simulated datasets from the Genetic Analysis Workshop 17 (GAW17) comprising unrelated individuals.
• Employed the support vector machine (SVM) algorithm to build disease risk prediction models.
• Compared model discrimination power using common variants versus models incorporating both common and rare variants.

Main Results:

• Empirical results demonstrate that rare variants contribute substantially to disease risk prediction.
• Models incorporating rare variants showed improved predictive capabilities compared to those relying solely on common variants.

Conclusions:

• Rare genetic variants possess appreciable effects on disease risk prediction.
• Incorporating rare variants enhances the quantification of genetic risk, leading to more accurate disease prediction models.