Barth syndrome (BTHS) is an X-linked genetic disease characterized by severe cardiovascular defects, skeletal muscle weakness and neutropenia. Defects in the protein tafazzin, which is encoded by the TAZ gene, have been identified to give rise to BTHS through impaired capability of remodeling the mitochondrial phospholipid cardiolipin (CL). Even though the genetic defects responsible for the disease are known, the links between specific TAZ mutation variants and mechanisms of BTHS pathogenicity haven't been thoroughly examined. The aim of this work was to present bioinformatics methods which would facilitate this type of research and, using these methods, examine tafazzin protein features, map all known benign and pathogenic tafazzin protein changes to its corresponding domains, identify the mutations that are found in evolutionarily conserved regions, and highlight potentially important domains and mutations for future in silico and in vivo research. The variants were compiled from the Barth Syndrome Foundation's Tafazzin Human Variants Database. Databases UniProt, ClinVar, OMIM and GnomAD were used in addition. Seven pathogenicity prediction softwares were used to analyze potential pathogenicity of missense variants based on protein structure and evolutionary conservation. A species alignment analysis was done using ClustalOmega. Our results show that tafazzin is not subject to binding CL via induced-fit, but that the interaction is rather based on conformational selection, indicating a potential conformational flexibility that might enable interactions with diverse phospholipids without significant structural changes caused by ligand binding. We hypothesize that tafazzin's conformational flexibility may be influenced by the surrounding microenvironment, including membrane composition and neighboring molecules. We demonstrate pathogenic mutational enrichment in five of the protein's domains and show that 34 out of 62 pathogenic missense variants are predicted to be detrimental to the protein's function. Through further refinement, we show that two domains of the protein, HX4D and MT1, are highlighted as significantly enriched in pathogenic variants. D74 and D75 mutations in the HX4D domain and R94 mutations in the MT1 domain, identified as pathogenic through our methods, suggest the likelihood of interactions between these domains due to the complementary charges of aspartate and arginine, hinting at a potential active site. Finally, we discuss the potential of studying these two domains and TAZ mutation variants in general by applying molecular dynamics simulations and using the model organism Saccharomyces cerevisiae.