This document summarizes a presentation on multi-trait modeling in polygenic scores. It discusses sparse regression models like LASSO to build polygenic risk scores (PRS) from large genetic datasets. It introduces multi-PRS models that combine disease PRS with biomarker PRS to improve disease prediction. It also presents genetic component-based PRS models called DeGAs-PRS that group genetic variants into components for enhanced interpretation. The presentation evaluates these multi-trait modeling approaches using large-scale UK Biobank data.