Blind prediction performance of RosettaAntibody 3.0: Grafting, relaxation, kinematic loop modeling, and full CDR optimization

By Brian D. Weitzner, Daisuke Kuroda, Nicholas A. Marze, Jianqing Xu, Jeffrey J. Gray. 2014

95 citations recorded by the source corpus.

Antibody Modeling Assessment II (AMA‐II) provided an opportunity to benchmark RosettaAntibody on a set of 11 unpublished antibody structures. RosettaAntibody produced accurate, physically realistic models, with all framework regions and 42 of the 55 non‐H3 CDR loops predicted to under an Ångström. The performance is notable when modeling H3 on a homology framework, where RosettaAntibody produced the best model among all participants for four of the 11 targets, two of which were predicted with sub‐Ångström accuracy. To improve RosettaAntibody, we pursued the causes of model errors. The most common limitation was template unavailability, underscoring the need for more antibody structures and/or better de novo loop methods. In some cases, better templates could have been found by considering residues outside of the CDRs. De novo CDR H3 modeling remains challenging at long loop lengths, but constraining the C‐terminal end of H3 to a kinked conformation allows near‐native conformations to be sampled more frequently. We also found that incorrect VL–VH orientations caused models with low H3 RMSDs to score poorly, suggesting that correct VL–VH orientations will improve discrimination between near‐native and incorrect conformations. These observations will guide the future development of RosettaAntibody. Proteins 2014; 82:1611–1623. © 2014 Wiley Periodicals, Inc.

Subfields: Immunology, Molecular Biology, Radiology, Nuclear Medicine and Imaging

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Immunology

In Molecular Biology

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Antibody modeling using the Prediction of ImmunoGlobulin Structure (PIGS) web server (2014, methodology): “Although recently we and others have developed new methods for the prediction of H3 loops with encouraging results 69–72 , the approach detailed above still appears to be sufficiently sensible from the standpoint of the accuracy of antibody models.”
Antibody Structure and Function: The Basis for Engineering Therapeutics (2019, background): “The participants included Accelrys, Inc. [46], Chemical Computer Group (CCG) [47], Schrödinger [48], Jeff Gray’s lab at John Hopkins University [49] Macromoltek [50], Astellas Pharma/Osaka University [51], and Prediction of ImmunoGlobulin Structure (PIGS) [52,53].”
AbPredict 2: A server for accurate and unstrained structure prediction of antibody variable domains (2018, methodology): “1B) — an improvement even relative to methods that used MolProbity as a model-selection criterion(Shirai et al., 2014; Weitzner et al., 2014).”
Modeling and docking antibody structures with Rosetta (2017, background): “When templates exist, errors in the FR and CDR L1–3, H1 and H2 loops are typically small (no greater than a 1-Å backbone RMSD to native) 5 .”
Improved prediction of antibody VL-VH orientation. (2016, background): “Failing to account for the VL–VH orientation during CDR or paratope structure prediction dramatically hinders the quality of the output models, and recent evaluation found the VL–VH orientation to be a limiting factor in antibody structure prediction (Weitzner et al., 2014).”
Accurate structure prediction of CDR H3 loops enabled by a novel structure-based C-terminal constraint (2016, methodology): “In AMA-II we found that producing low-RMSD models required filtering out nonkinked H3 conformations.”