Abstract Cystic fibrosis (CF) is a lethal genetic disease that currently affects ~100,000 people worldwide. CF is caused by a spectrum of loss-of-function mutations that compromise the biogenesis and/ or function of the cystic fibrosis transmembrane conductance regulator (CFTR) ion channel, most of which enhance its misfolding and degradation. Recent drug discovery efforts have yielded a suite of approved small molecule “correctors” that enhance the expression of misfolded CFTR variants and “potentiators” that restore conductance to CFTR variants with defective gating. Combinations of these molecules have recently revolutionized the treatment of the ~90% of CF patients bearing at least one copy of the well-studied ΔF508 CFTR variant, which is highly penetrant among Caucasians. However, the efficacy of current combinatorial therapies varies widely among the ~10% of patients bearing diverse combinations of rare, uncharacterized CF variants with divergent pharmacological properties (“theratype”). Efforts to expand the labels of current therapeutics and maximize the number of treatable CF genotypes, in particular amongst non-white populations, are constrained by the large number of CF variants and the limited throughput of current methods. Identifying rare CF variants that respond to therapeutic cocktails is likely to become even more challenging as new correctors and/ or potentiators gain approval. Addressing this challenge requires new techniques that enable efficient biochemical and/ or pharmacological profiling of rare CF variants. In the following, we propose to address this challenge with a unique fusion of emerging genetic, biochemical, and computational methods. We show how deep mutational scanning (DMS) can be used to compare the effects of correctors on the expression of hundreds of variants in parallel. Our preliminary findings provide an unprecedented glimpse of the divergent theratypes of CF variants while identifying numerous variants with unique biochemical and/ or pharmacological properties. We first propose to expand on these investigations in order to measure the response of the complete set of CFTR2 missense variants to a panel of structurally diverse corrector molecules. We will then characterize the interactomes of variants with distinct corrector responses to identify CFTR interactions that antagonize the effects of these small molecules. We will then fuse CRISPR/ Cas9 technology with DMS to determine how these interactions impact the spectrum of CF variant theratypes. Using state-of-the-art structural modeling approaches, we will then identify structural defects in the CFTR protein that are associated with the formation of antagonistic interactions and deviations in CFTR variant theratype. We will then utilize machine learning to classify CF variants based on their observed pharmacological properties. Finally, we will assess the effects of approved correctors on the functional properties of previously uncharacterized variants u...