Assessment of PRediction of Interaction) is a community wide experiment
designed to test methods to predict the structure of macromolecular complexes
based on the known structure of their components (http://www.ebi.ac.uk/msd-srv/capri/). The experiment,
inspired by CASP, was started in 2001. It has now completed 47 prediction rounds,
each with 25-40 participating groups, and a total of 162 targets. A CAPRI target
is an unpublished crystal, NMR or high resolution cryo-EM structure of a
protein-protein complex, communicated on a confidential basis by their authors
to the CAPRI management. In recent years, Protein-peptide, protein-RNA
complexes, and designed complexes for which affinity had to be predicted along
with structure, have also been CAPRI targets.
CAPRI has proven to be
very useful to the community of computational biologists. Like CASP, it has
greatly fostered the development of novel methods, like CASP, its continuation
depends on the willingness of structural biologists to offer unpublished
experimental structures as targets. But the realm of methods for predicting the
structures of multi-component macromolecular complexes has significantly
expanded in recent years. With the exploding landscape of available protein
sequences it has become possible to model the structures of individual proteins
with increasing accuracy from sequence information alone and these models could
soon be used as building blocs to assemble larger macromolecular complexes,
using computation methods and additional experimental clues. As an important
step towards achieving this goal CAPRI has conducted several joint prediction
experiments with CASP, in addition to the standard CAPRI rounds. The latest
joint experiment was completed in the summer of 2018, with results to be
In this webinar, we will
briefly trace the history of CAPRI and protein docking. We will describe how
CAPRI is organized, who participates, and how it has helped build a very
dynamic, and highly collaborative community, passionate about developing better
methods, training newcomers to the field, and readily exchanging programs and
tools. We will end by outlining CAPRI’s hopes and plans for the future, and
present proposals for potential collaborations with BioExcel.
A glimpse of what the
future may hold for these approaches was obtained in the recent prediction
experiment run jointly by the CASP and CAPRI community, with results presented
at the CASP11-meeting in Dec. 2014. In
this experiment members of the CASP community predicted the structures of the
individual protein components, whereas members of the CAPRI community predicted
the structures of the complexes formed by these components using classical
docking methods or extrapolating from the structures of related complexes
already deposited in the PDB. While this first joint experiment focused
primarily on complexes comprising identical subunits, the experiment
demonstrated beyond any doubt that integrations of different modeling
techniques with information from various sources will the prevailing approach
in the future.
Shoshana Wodak and Marc Lensink
Shoshana J. Wodak, earned her PhD at Columbia University, New York, under the supervision of Cyrus Levinthal, who was developing what were then the first and most powerful computational and graphics tools for modelling proteins and protein interactions. Importing and implementing many of these tools to Belgium and France in the late seventies (with crucial support from CECAM (Centre pour le Calcul Atomique et Moléculaire) in Paris, and EMBO), she developed, in collaboration with Joel Janin, the first docking algorithms for the prediction of protein-protein interactions and one of the first procedures for defining structural domains from the atomic coordinates of proteins. With her team at the Free University of Brussels she then used molecular simulations and bioinformatics approaches to investigate the role of local interactions in stabilizing the native state of proteins, protein structure prediction, protein folding, and fold recognition. Together with her teams at the Free University of Brussels, the European Bioinformatics Institute in Hinxton, UK, and at the Hospital for Sick Children in Toronto, they also developed efficient procedures for computational protein design, for simulating protein interactions and conformational changes, and for analysing protein interactions networks and cellular pathways.
Between 1986-1993, Dr. Wodak
was the Scientific Director of the protein engineering team at Plant Genetic
Systems, Belgium (now part of Bayer CropScience). Among their success stories
then, were the engineering more thermo-stable versions of the xylose (glucose)
isomerase enzyme, changing its metal specificity, and pH profile.
Dr. Wodak founded and
co-directed the Centre for Structural Biology and Bioinformatics and headed a
Master’s program in Bioinformatics. She has been a member of the European
Molecular Biology Organization (EMBO) since 1990. She held a Tier 1 Canada
Research Chair in Computational Biology and Bioinformatics from 2005-2012, and
was elected an ISCB fellow in 2016. She is a member of F1000 since 2017.
Dr. Wodak has been a member
of numerous expert panels and Advisory Committees in Europe (ERC, Horizon
2020), US (DOE) and Canada (CIHR), and is on the Editorial Boards of several
journals her my field.
Since 2001 Dr. Wodak serves
on the Management Committee of CAPRI (Critical Assessment of Predicted
Interactions), a community-wide international initiative for fostering the
development of methods and algorithms for the prediction of protein interactions
and complexes, and is currently coordinating this effort.
Dr. Wodak is currently a
Visiting Group Leader at the VIB-VUB Structural Biology research Centre, at the
Flemish Free University of Brussels, Belgium.
Marc Lensink obtained his PhD in Mathematics and Natural Sciences at the University of Groningen in 2002. After post-doctoral fellowships in Finland, Belgium and France he was appointed a CNRS permanent researcher in 2011, affiliated with the University of Lille. Since 2015 he is team leader of the Computational Molecular Systems Biology group at the Institute for Structural and Functional Glycobiology.
Marc Lensink is an expert in biomolecular simulations with
an interest in protein-protein, protein-lipid and protein-carbohydrate
interactions. He has authored 54 original research papers, which have been
cited a total of 1600 times (H-index of 20).
Marc Lensink has been responsible for the evaluation of
CAPRI submissions since 2004. He is member of the CAPRI Management Committee,
CAPRI on twitter: @CAPRIdock
Register for webinar
Title: Prediction of protein-protein interactions in CAPRI: an increasingly integrative approach
Date: 27th June, 2019 Time: 14:00 BST / 15:00 CEST