success of antibody-based drugs has spurred the refinement of methods to
engineer more efficacious, safer and developable therapeutic antibodies. An
important component of the antibody engineering process is the availability of
high quality three-dimensional (3D) structures of the potential antibody-based
drug. Such information is useful to guide humanization, assist the design of
libraries of mutants for optimization of binding properties such as kon and
koff, and engineer variants with improved solubility, less aggregation and
structure determination by x-ray crystallography is relatively robust, but it
is resource intensive and in some cases the turnaround time is unpredictable,
relative to computational modeling methods. Thus, the question is how an
antibody 3D model compares to an experimentally determined high resolution
structure of the antibody and therefore how reliable predictions and designs
made based on models turn out to be.
address this question we have conducted two Antibody Modeling Assessments (AMA’s).
For the first assessment (AMA-I) nine unpublished
high-resolution x-ray Fab crystal structures covering a wide range of
antigen-binding site conformations were used as benchmark to compare Fv models
generated by four structure prediction methodologies. For the second assessment
a more diverse and larger set (eleven unpublished structures) was used as
benchmark to compare seven structure prediction methodologies.
webserver compiles information of the assessments, including the benchmark
structures, models generated by the participants and highlights of AMA's, together
with resources and literature on Antibody modeling.