Homology modelling

Three-dimensional (3D) models of SET domains identified in the C. parvum genome (CpSETs) were built with the automated comparative modelling program Swiss Model Interactive Workspace (https://swissmodel.expasy.org/interactive) using as homologous protein templates, highly resolved X-ray crystal structures of human SET1 (MLL1) Protein Data Bank (PDB) code: 5F6 L; X-ray resolution 1.90 Å, SET2 (SETD2) (6J9J; 1.78 Å), and SET8 (SETD8) (5TEG; 1.30 Å). For each CpSET model developed, the quality of the structure was evaluated by the MolProbity web server [³⁶]. The MolProbity score is a combination of the clash score, rotamer and geometric parameters, and the Ramachandran evaluations into a single score [³⁷]. Lower MolProbity scores are better, meaning good quality structures. The MolProbity server reports also a percentile relative to the score distribution for crystal structures near the resolution of the submitted structure. In case of a modelled structure, the distribution is established covering all resolutions (range of 0Å-99Å). Distance matrix alignment (Dali) server [³⁸] was used to perform pairwise structural alignment between the template and the newly generated CpSET models. The secondary structures were assigned using DSSP algorithm and the ChimeraX software was used to visualize the superimposition of templates and CpSET models [³⁹].