Due to the spread of coronavirus this event has been rescheduled for the Fall of 2020. New information will be posted as soon as it becomes available.
** Goal: * Provide overview of state-of-the-art approaches for modeling protein interactions. * Understand and master concepts of integrative modeling of multiprotein systems within their biological setting. * Make modelers familiar with relevant experimental approaches for validation, extension and integration of models. **
Protein interactions are the critical building blocks of cell functionality, hence understanding structure of protein complexes and their association pathways is the key to understanding mechanisms of biology. Despite progress in the experimental determination and computational prediction of individual protein structures, obtaining structures of protein complexes, especially those participating in signal transduction as well as flexible multiprotein assemblies, remains challenging. Hence, computational modeling of protein interactions, i.e., the structure of protein complexes (or docking), is of great importance.
Promising recent developments have allowed significant progress in the field. Not only have methods for global domain–domain protein docking achieved maturity [1,5], but also flexible global domain-peptide (unstructured proteins) docking methods are now available to extend the span of different interactions that can today be modeled at atomic level . Major advances in cryo electron microscopy have brought into reach structural pictures of multiprotein complexes and machineries of increased complexity. Due to their often low resolution, they have spurred the development of a wealth of new algorithmic approaches for the effective fitting of individual proteins into the low-resolution density maps. Most recently, data-driven algorithms equipped with artificial intelligence have made a big leap forward in computational structural biology, leading to the claim at last CASP that “the problem of protein structure prediction from its sequence has been essentially solved” . This was made possible thanks to the massive increase in natural protein sequences and the accompanying rapid progress in multiple experimental and computational disciplines, most notably sequence analysis, coevolution models  and deep neural networks.
Despite progress in the modeling of monomer structures and specific types of protein interactions, several main bottlenecks need to be overcome towards the realistic and comprehensive modeling of protein interactions in a cell. To extend the scope of modeled interactions we need to (1) master docking of structural models (instead of solved structures), and (2) improve the definition and application of coevolutionary constraints to protein interfaces (as used successfully for modeling of monomers); special tools need to be developed to model (3) flexible interactions involving intrinsically unstructured regions (e.g. conformational heterogeneity and fuzzy interactions), as well as (4) aggregates [6,7]. For them to be relevant, at least part of these models will need to consider the biological setting (e.g. the crowded environment) .
This CECAM workshop will bring together computational biophysicists, bioinformaticians, as well as experimentalists, to critically assess the state of the field and lay out new plans on how to address and tackle the problems encountered with complex protein interactions within their natural, often multi-partner setting and will focus on a deeper integration of experiments with simulations.
An explicit goal of the workshop will be to consolidate the eastern-European teams and introduce them to the challenges in the field. We expect about one third of the invited speakers to be from Eastern Europe (UA, PL, LT, EE, HU, RU).
- Bring together modeling experts from different subcommunities to describe the state of the art and challenges in the modeling of protein interactions
- Discuss future directions in the field and determine focus research axes
- Identify complementary collaborations among experts of different approaches, for general methods improvement, as well as for specific applications to biological systems of interest
- Tighten the connection between experiment and modelling
- Introduce frontiers of the field to the eastern-European teams