Interpreting missense variants in highly conserved, paralog-rich gene families remains a major barrier to understanding protein function and advancing precision medicine. Here, we combine comprehensive mutagenesis, high-content live-cell imaging, and automated quantitative analysis to generate a complete functional atlas of all 2,683 single-nucleotide coding variants in human α-tubulin TUBA1A. Systematic profiling of microtubule assembly revealed distinct classes of mutations that disrupt folding, chaperone engagement, and protofilament geometry, defining structural constraints that govern tubulin function. This approach complements conservation-based predictors and enables functional reinterpretation of disease-associated variants. Molecular dynamics simulations further revealed how local perturbations in GTP (guanosine triphosphate) binding, dimer contacts, and lateral interfaces propagate to alter filament architecture. Integrating these datasets produced a predictive framework that generalizes across tubulin isotypes and species, enabling accurate inference of variant effects. This work establishes a scalable, high-resolution strategy for functional annotation of conserved proteins and provides mechanistic insight into cytoskeletal assembly and flexibility.
Xu et al. (Fri,) studied this question.