Second antibody modeling assessment (AMA‐II)

Abstract

To assess the state of the art in antibody 3D modeling, 11 unpublished high-resolution x-ray Fab crystal structures from diverse species and covering a wide range of antigen-binding site conformations were used as a benchmark to compare Fv models generated by seven structure prediction methodologies. The participants included: Accerlys Inc, Chemical Computer Group (CCG), Schrodinger, Jeff Gray's lab at John Hopkins University, Macromoltek, Astellas Pharma/Osaka University and Prediction of ImmunoGlobulin Structure (PIGS). The sequences of benchmark structures were submitted to the modelers and PIGS, and a set of models were generated for each structure. We provide here an overview of the organization, participants and main results of this second antibody modeling assessment (AMA-II). Also, we compare the results with the first antibody assessment published in this journal (Almagro et al., 2011;79:3050).

Homology modeling and structure-based design improve hydrophobic interaction chromatography behavior of integrin binding antibodies

“…The accuracy of homology modeling is highly dependent on the sequence similarity of the structural template(s) used, and for mAb variable domains, the framework modeling is generally successful (backbone RMSD < 1), as are the LC CDRs and HC CDRs 1 and 2. The CDR-H3 accuracy varies in homology modeling and this, 26,27 as well as sidechain conformational variation, can explain why the MD averaging improves experimental predictions. The conformational ensemble produced by LowModeMD likely smoothes out sensitivity in homology modeling, reducing error, especially when template quality is poor.…”

A Combination of Structural and Empirical Analyses Delineates the Key Contacts Mediating Stability and Affinity Increases in an Optimized Biotherapeutic Single-chain Fv (scFv)

“…This study shows that the low stability of scFvs built on theoretically stable v-domain scaffolds (14), whether fully human or humanized, can be driven by as little as a single germline side chain clashing with the V H -CDR3. As V H -CDR3 structures are still problematic to model, particularly for human antibodies, which are poorly represented in the PDB database (36), such clashes in CDRs can be rapidly identified and ameliorated by CDR mutagenesis, and subsequently rationalized by structural biology.…”

Prediction of methionine oxidation risk in monoclonal antibodies using a machine learning method

“…Moreover, these mispredicted cases in the CDR-H3 region are also located on very long loops (length given in brackets): M97 (11), M100b (13), M100h (16), M100i (19) and M100* (20). This suggested that the failure of the model in these cases was most likely due to uncertainty in the CDR-H3 modeling, 32,33 which would affect the features extracted from these structures.…”

“…This becomes especially important for long CDR-H3 loops, where even state-of-the art models still lack reliability. 32,33 It is also worth noting that predictive power is directly influenced by dataset size; therefore, the current model can be further improved (especially to minimize the number of false positives), with increased dataset size (specifically, liable Met) in the future. Some studies have shown that residues that fall outside of the traditionally defined CDRs can also be important to antigen binding, 37 which suggests that molecular assessment studies may need to be further extended to these residues.…”