A computational mutagenesis methodology utilizing a four-body, knowledge-based, statistical contact potential is applied toward globally quantifying relative structural changes (residual scores) in bacteriophage f1 gene V protein (GVP) due to single amino acid residue substitutions. We show that these residual scores correlate well with experimentally measured relative changes in protein function caused by the mutations. For each mutant, the approach also yields local measures of environmental perturbation occurring at every residue position (residual profile) in the protein. Implementation of the random forest algorithm, utilizing experimental GVP mutants whose feature vector components include environmental changes at the mutated position and at six nearest neighbors, correctly classifies mutants based on function with up to 72% accuracy while achieving 0.77 area under the receiver operating characteristic curve and a 0.42 correlation coefficient. An optimally trained random forest model is subsequently used to infer function for all remaining unexplored GVP mutants.