Project

Computational Biophysics; G-protein coupled receptors; QM/MM Car-Parrinello/TDDFT; Biological Signalling; multiscale modelling; Biomimetic Design; Biochemistry; Computational Chemistry

Lead
Approches de simulation à plusieurs échelles appliquées à la signalisation biologique et au design bio-inspirée
Lay summary
Par le truchement de millions d’années d’évolution, la nature a développé des processus aptes à maintenir des écosystèmes hautement efficaces et stables. Il n’est donc pas surprenant d’y puiser notre inspiration afin de développer des stratégies durables visant à produire des ressources matérielles et énergétiques dont la population mondiale grandissante a besoin. Cependant, une compréhension complète des fonctionnements internes de processus biologiques est très fréquemment une tâche ardue, vu que ce sont des événements complexes pouvant impliquer plusieurs ordres de grandeur tant en temps qu’en dimension et que les techniques expérimentales testant leur mode d’action ne fournissent souvent que des informations partielles ou indirectes. La simulation par ordinateur donne de nos jours un aperçu direct de ces mécanismes biomoléculaires, mais la combinaison d’une taille étendue, d’un large espace des configurations disponibles et d’une haute précision, nécessaire afin de décrire les différences d’énergie faible de l’ordre de kT, reste un défi. Dans cette recherche, nous proposons de (1) Repousser les limites courantes des simulations biologiques à plusieurs échelles, grâce au développement d’une approche hybride généralisée QM/(QM)/MM qui sera capable de résoudre simultanément l’exigence liée à la précision nécessaire en taille et en temps; (2) Développer et implémenter des techniques d’analyses de données efficaces et systématiques, basées sur la sélection de caractéristiques et l’inférence causale, qui nous permettront d’identifier les facteurs cruciaux responsables d’une transformation biologique donnée; (3) Utiliser cette information afin de guider la conception de systèmes bio-inspirés et (4) Optimiser les systèmes biomimétiques afin d’obtenir des propriétés sur mesure en utilisant l’exploration efficace d’un espace chimique grâce à des algorithmes évolutionnistes.

Direct link to Lay Summary

Last update: 08.04.2013

Active participation

Title	Year

Number	Title	Start	Funding scheme

The long-term survival of our society may critically depend on our ability to find and implement sustainable strategies that are able support our ever-increasing demands for material and energy resources while maintaining an intact environment. Through millions of years of evolution, Nature has managed to develop processes that are able to maintain highly efficient and stable ecosystems. It is therefore not surprising to seek inspiration from Nature to develop sustainable strategies to produce the material and energy resources needed for the world’s growing population. The first step towards the development of biomimetic systems is to understand how Nature ‘is doing it’. However, a comprehensive understanding of the inner workings of biological processes is often a very difficult task since they tend to be notoriously complex events that can involve several orders in length and time scales and experimental techniques probing their mode of action can often provide only indirect and partial information. Computer simulations can nowadays give direct insights into biomolecular mechanisms but the combined challenge of extended size, large available configurational space and high accuracy required to describe small energy differences of the order of kT remains a challenge.Here we propose to (1) push forward the current frontiers of multiscale biological simulations through the development of a generalized QM/(QM)/MM hybrid approach that will be able to address the combined challenge of size-time-accuracy requirements simultaneously; (2) Develop and implement systematic and efficient data analysis techniques based on feature selection and causality inference that allow us to identify the crucial factors that are responsible for a given biological transformation; (3) use this information to guide the design of bioinspired systems and (4) Optimize biomimetic systems for tailored properties via an efficient exploration of chemical space based on evolutionary algorithms.In continuation of our current SNF grant 200020-120082, we will apply methods (1) and (2) to two electronically triggered biological signaling cascades, electron-transfer induced DNA repair and signal transduction in G protein - coupled receptors. Since these biological signaling processes span many orders of length and time scales, from the electronic/atomic level to the mesoscopic/microscopic domain with accompanying time windows that range from 10-15 to 10-3s, these are two problems par excellence to test the performance of new multiscale methods. Furthermore, dysregulation of signaling cascades lie at the heart of a large variety of diseases such as cancer and neurodegeneration and a better understanding of the underlying molecular mechanisms of these processes might ultimately assist in the development of efficacious therapeutic approaches against these diseases.The newly developed methods (3) and (4), we will apply for the computational design of a biomimetic system for CO2 fixation and to the optimization of biomimetic porphyrin dyes in dye-sensitized solar cells (DSSCs). Both projects will be performed in close collaboration with experimental groups. The requested resources for this project are 4 PhD positions for 3 years. With this research, we expect to (i) make major advances in the development of computer simulation methods for biological systems; (ii) gain fundamental understanding of the molecular mechanisms involved in DNA repair and GPCR signaling; and (iii) contribute to the development of biomimetic compounds for CO2 fixation and DSSCs with higher efficiencies.