AlphaMissense is a new tool for predicting the pathogenicity of missense variants in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The study evaluates its performance against CFTR variants from the CFTR2.org database. AM predicted high pathogenicity for CFTR residues, resulting in a high false positive rate and fair classification performance. AM pathogenicity scores correlated modestly with clinical metrics such as sweat chloride levels, pancreatic insufficiency, and Pseudomonas aeruginosa infection rates. However, AM scores correlated well with CFTR channel function in vitro, indicating that the dual structure and evolutionary training approach learns important functional information. AM performed well in predicting the pathogenicity of severe CF-causing variants but showed limited utility in predicting pharmacological responses, such as theratype. AM scores correlated modestly with trafficking and folding competency in vitro but well with channel function. AM predictions offered limited utility in informing on the pharmacological response of CF variants. The study highlights the need for new approaches to differentiate the biochemical and pharmacological properties of CFTR variants to refine the targeting of emerging precision CF therapeutics. The results suggest that AM may offer capabilities in predicting the pathogenicity of emerging variants but proved less useful for theratyping variants. The study also shows that AM scores correlated modestly with clinical data and performed poorly on VVCCs and VUSs. AM scores were found to be closely aligned with pathogenicity but could not differentiate between variants that compromise expression versus function. The study concludes that AlphaMissense has the potential to aid in the classification of rare and emerging variants identified during genetic screening. However, it does not appear useful for CFTR theratype predictions. The study also highlights the need for further research to improve the accuracy of pathogenicity predictions for CFTR variants.AlphaMissense is a new tool for predicting the pathogenicity of missense variants in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The study evaluates its performance against CFTR variants from the CFTR2.org database. AM predicted high pathogenicity for CFTR residues, resulting in a high false positive rate and fair classification performance. AM pathogenicity scores correlated modestly with clinical metrics such as sweat chloride levels, pancreatic insufficiency, and Pseudomonas aeruginosa infection rates. However, AM scores correlated well with CFTR channel function in vitro, indicating that the dual structure and evolutionary training approach learns important functional information. AM performed well in predicting the pathogenicity of severe CF-causing variants but showed limited utility in predicting pharmacological responses, such as theratype. AM scores correlated modestly with trafficking and folding competency in vitro but well with channel function. AM predictions offered limited utility in informing on the pharmacological response of CF variants. The study highlights the need for new approaches to differentiate the biochemical and pharmacological properties of CFTR variants to refine the targeting of emerging precision CF therapeutics. The results suggest that AM may offer capabilities in predicting the pathogenicity of emerging variants but proved less useful for theratyping variants. The study also shows that AM scores correlated modestly with clinical data and performed poorly on VVCCs and VUSs. AM scores were found to be closely aligned with pathogenicity but could not differentiate between variants that compromise expression versus function. The study concludes that AlphaMissense has the potential to aid in the classification of rare and emerging variants identified during genetic screening. However, it does not appear useful for CFTR theratype predictions. The study also highlights the need for further research to improve the accuracy of pathogenicity predictions for CFTR variants.