(The english side of ABC&T website is in process of being updated, please rather refer to french side if possible)

Welcome on our site !

The Association of Theoretical and Computational Biophysics exists since the end of 2006 and was created on the initiative of university researchers being interested in life sciences and using multidisciplinary approaches (molecular biology, physics, quantum chemistry, mathematics).

In the current context of the French research, an association also allows to give a framework of an innovative kind for the exercise of the vocations of fundamental research.

A few definitions

Biophysics is a field where the study of the structures and the functions which characterize life is made under the light of the knowledge of physics, as the name indicates. The borders between disciplines, more compartmentalized formerly, widen more and more to give birth to new domains, interdisciplinary, such as bioinformatics, mecatronics, cognitive sciences, etc. But sciences can only get involved inextricably in their applications and for the deep understanding of the universe which contains us: physical chemistry, quantum chemistry, biophysics, the mathematics being the cement of all. In our case we are interested not in macroscopic biophysics, but in biophysics at the molecular or supramolecular scale.

Computational : by calculation performed on a computer. The word 'computationnel' in French is mentioned in no dictionary of the French language, but is widely used by the experts of the field or in English. It derives from the verb 'computer' (' computare', to calculate in Latin), used in the XVIth century to determine a date or in the XIXth century with the meaning of to estimate. The English people thus borrowed us this word to make 'to compute' (calculate on computer) and 'computer'! Computational biophysics thus uses the computer and (bio)informatics as tools of investigation, both to make much too complex calculations to be hand-made, and to model, i.e. simulate a phenomenon on the basis of a model supported by precise preliminary experimental knowledge. The modelling allows to predict a certain number of properties, structural or functional: the three-dimensional structure of a protein, the mechanism of a reaction, etc. The computer allows to test theories and to produce 'observations' sometimes impossible to obtain in the reality.

Theoretical: sometimes abstruse but not equal to misty! Theory tries to formulate the framework describing best the results and the experimental observations, with a minimal number of hypotheses. Theory progresses and becomes refined in a constant parallel with experiment and in the most perfect adequacy possible. The data are sometimes insufficient or imprecise, theory allows to confirm or invalidate, to correct, to explain. It can open the way but also sometimes engage on sidings, even create a conceptual environment whose dogmatism can prevent from evolving rightly. The critical test is always the confrontation with experimental data.

Computer simulation

The computer constitutes today a tool of fundamental research in itself, with its assets and its limitations, in the same way as a technique of experimental investigation with its intrinsic specificities. It opens multiple windows of observation of the real world, by totally discovering certain of its aspects, by clearing up others. There is not anymore only the lab work surface or the final equation on the paper to shout eurêka, there is the screen of the computer! One should be conscious of it.

The power of computers or supercomputers (networks or clusters) exceeds what we could imagine not long ago. The gigaflops (one billion floating-point operations per second) is the measure for a personal computer, and one approaches the teraflops (a trilliard flops) with a Sony PlayStation 3 console. The computer BlueGene/L from IBM reaches 367 teraflops with 131072 processors, just behind the Japanese RIKEN MDGRAPE-3 which is getting close to the petaflops (a million billions flops)! This last one is dedicated to molecular dynamics simulations.

But of course power is not everything, softwares are numerous and sophisticated today. In the field of the calculation of the dynamics of biological molecules one can quote AMBER, GROMACS, CHARM++, NAMD, in that of quantum chemistry: GAMESS, GAUSSIAN, HYPERCHEM, of programming: FORTRAN, JAVA, PERL, of formal calculation: MAPLE, MATHEMATICA, MATLAB, SCILAB, not to mention editing, visualization… Enormous software improvements allowed to integrate in these environments the most recent theories, and to be able to apply them to systems which can contain up to several hundreds of thousands atoms. GAUSSIAN, a set of integrated tools, was initially developed by Sir John Pople, who was rewarded for that by the Nobel prize of Chemistry in 1998.

What kind of things are simulated? From quantum chromodynamics to the evolution of the galaxies or the dynamics of the universe, a great deal of aspects are simulated today: turbulence in plasmas, meteorological phenomena, embryogenesis, behavior of social insects, production of the sound of instruments, growth of plants, geometry of biomolecular interactions, fluids flows (air, blood), evolution of financial markets, walk, vision, hepatic surgery, self-organized systems (cellular automata), etc. An important way of computer research consists in molecular modelling, namely for example figure a protein in its aqueous (such as the inside of a cell) or lipidic (a biological membrane, the wall of any cell) environment, or a small part of the DNA (support of the genetic information), and simulate the evolution of this system or study a given reaction within it. To reach a good adequacy between the results of such studies, named in silico, and the experimental data, more and more works mix the classical (Newton equations) with the quantum (exact description of the interactions notably between the electronic clouds of atoms). Calculation times are very variable, dependent on the complexity and the size of the studied system, and of course on the type of calculation and the power of calculation at disposal. A classical molecular dynamics trajectory of a protein in its lipidic or aqueous environment can take several months to one year to compute, describing an interval of time of a few nanoseconds (billionths of second). To add it a quantum description blows up calculation times. A calculation at the quantum level on a snapshot of such a configuration can be realized on the other hand in a time of the order of the hour up to several days.

Our works

(See ONGOING PROJECTS (yet in french) for more precisions)

Within our association we study diverse biological mechanisms of fundamental importance:

- A reaction within a protein, the ANT, inserted into the inner membrane of an organelle called mitochondrion (the factory producing the chemical energy of cells). The ANT undergoes, through a reaction of oxidation between two of its aminoacids (formation of a cystine bridge), mediated by a reaction with a small molecule called glutathione, a change in conformation (its structure in space). This is going to lead to a spate of events ending in the suicide of the cell, called apoptosis (programmed cell death). To better understand the functioning of this protein at the origin of apoptosis will allow to target it by drug molecules so as to provoke only the death of cancer cells.

- The reaction between a highly carcinogenic molecule, benzopyrene (in its epoxidized form), and a base of the DNA. Benzopyrene is present in the cigarette smoke or the exhaust gases of diesel engines, and by reacting with a base of the DNA, guanine, is going to lead to lung cancer by inducing mutations in certain genes. Researchers demonstrated that when the base next to guanine, cytosine, possesses a methyl group (CH₃), the guanine is going to fix much better the carcinogenic molecule. It is fundamental to understand indeed the origin of this difference of reactivity. The application of the tools of quantum chemistry for the study of cancer is a fundamentally innovative approach in the world of research. We have this experience and the elucidation of this mechanism will allow to better address the origin of a lung cancer.

- The mecanistics of ionic channels. An ionic channel is a protein inserted into the biological membrane of a cell. It allows to transport from one side of the membrane to the other (for example from the extracellular medium towards the intracellular one) an ion in a selective and extremely effective way. We more specifically study the potassium (K⁺) channel, voltage-gated because its functioning depends on the difference of potential existing between both sides of the membrane. A model of this channel, the structure of which was resolved by Roderick MacKinnon (Nobel prize of Chemistry 2003), is the one of a bacterium, Streptomyces lividans, KcsA. In certain neurodegenerative pathologies such as Alzheimer's disease, the neuronal loss ensuing from the apoptosis of neurones is important because it can concern up to 50 % of neurones in certain brain areas. Now it was shown that quite as for the calcium, a massive exodus of potassium can provoke the phenomenon of programmed cell death. An action on potassium channels can thus constitute a therapeutic strategy to slow down neuronal degeneration.

- Protein folding is the Holy Grail of modern biology. As the genetic code (a triplet of bases of the DNA codes for an aminoacid), a code hides itself in the sequence which defines the succession of aminoacids (the elementary bricks) of proteins. Of this sequence, said primary (I), is going to depend the spatial structure of the protein when it will be folded on itself, starting from an initially disordered conformation. The precise and unique final three-dimensional conformation will correspond to the active, functional protein. We get down to clarify the role of the partial electronic charges (fractions of the charge of an electron, dependent on the nature of the atom and on the geometry of the aminoacids chain) for the folding dynamics, as well as to precise a thermodynamic model of folding/unfolding for a particular protein. Any step realized in the direction of a better understanding of the mechanism of folding of proteins is of fundamental importance. Indeed, the only way to get to the 3D structure of a protein is to crystallize it (what can sometimes take years or turn out to be impossible, particularly for membrane proteins) and characterize it by X-rays diffraction, or to perform a study by Nuclear Magnetic Resonance (NMR) in solution, what is long and difficult. The current models do not allow sharp prediction. The human genome was deciphered, but one is far from knowing the structures of proteins coded by all the genes, and thus their function. Furthermore, it is easy to synthesize a given protein sequence, and the knowledge of the link sequence I/3D structure would allow to design numerous medicines.