Since the rise of the omics era, ever-improving sequencing technologies have drastically expanded the horizons of genomics, leading to the discovery of tens of millions of DNA variants across human populations and the identification of tens of thousands of disease-associated mutations. Nonetheless, the majority of these variants remained functionally uncharacterized and thus far, the genomic revolution has seldom translated into biological and therapeutic applications. It has become increasingly clear that the functional complexity of living systems results from multiple molecular layers beyond the genome and that, integration of multiple omics data (such as genomics, transcriptomics, and proteomics/interactomics) holds the promise to unlock novel and actionable insights into health and disease. In this dissertation, I present several integrative omics frameworks to aid in the identification and interpretation of human disease mutations. In Chapter 2, I describe a cloud-based online platform that automates large-scale variant prioritization, offering a centralized workflow with high-level customization and a comprehensive collection of bioinformatics tools and omics data libraries. In Chapter 3, I present an experimental-computational integrated, interactome perturbation framework that for the first time, prioritizes functional missense mutations for human disease on a proteomic scale. In Chapters 4 and 5, I built upon the interactome perturbation framework with multiple modalities of omics data to further accelerate the discovery of potential causative changes in disease, and/or the actionable targets, that can be finally translated into rational design of new therapeutic strategies. In sum, the theme is two-fold: from the myriad number of DNA variants, (1) to identify which ones contributed to the disease and (2) to interpret how they have contributed.