Reproduction (biologie) | EPFL Graph Search

En biologie, la reproduction est un processus biologique qui permet la production de nouveaux organismes d'une espèce à partir d'individus préexistants de cette espèce. Avec la nutrition, c'est une des grandes fonctions partagées par tous les organismes vivants, assurant la continuité de l'espèce qui, sans reproduction, s'éteint. La reproduction peut être couplée à un système de dispersion dans l'espace. Il s'agit de systèmes permettant de coloniser de nouveaux biotopes, et d'augmenter les chances de survie des espèces. Pour certains auteurs, le terme « reproduction » serait à réserver à la seule reproduction sexuée. Or, dans la littérature, le terme recouvre généralement à la fois la reproduction sexuée et la reproduction asexuée. La reproduction, souvent considérée comme une évidence au sein du vivant, pose en fait des questions évolutives multiples. La fréquence relative des évènements de reproduction sexuée et asexuée varie en effet se

On isotopically selective collisional energy transfer and infrared multiple photon absorption by vibrationally highly excited molecules

Richard Bossart

This study focuses on an interesting and important phenomenon that was employed in a new approach to laser isotope separation that has been recently proposed and developed in our laboratory for highly selective separation of carbon isotopes. This approach consists of pre-exciting CF3H molecules with the desired isotope to a low vibrational overtone of the CH-stretch with a subsequent selective infrared multiple photon dissociation (IRMPD) of only the pre-excited molecules by a CO2-laser pulse. Significant isotopic shifts of the employed CH overtone bands already allow high selectivity at the pre-excitation step. This selectivity, however, can be further greatly increased by increasing the pressure of the sample gas and/or the delay between the two laser pulses; that is, by increasing the number of molecular collisions during the process. At first glance, the observed effect contradicts the general expectations that the isotopic selectivity of such a process should drop with an increase in the number of collisions because of vibrational energy transfer between different isotopic species. In this work we have studied this phenomenon and found its physical origins. We propose the contribution of two different mechanisms to the observed enhancement of isotopic selectivity by collisions. First, the vibrational collisional relaxation itself is isotopically selective, that is vibrationally excited 12CF3H relax on the bath of cold, 12CF3H, molecules faster than the excited 13CF3H do on the same bath. The primary reason for such a selectivity could be a significant isotopic shift of the vibration (CF-stretch, we believe) that mediates the energy transfer. A second mechanism arises from the difference between the IRMPD probabilities of the two isotopic species by the CO2 laser tuned to a particular wavelength that enhances the dissociation yield of the targeted, carbon-13, species. As collisional deactivation of both species increases the number of photons they have to absorb to be dissociated, this difference in dissociation probability increases as well, yielding an additional isotopic selectivity of the process. We perform a set of experiments and numerical simulations to investigate these two mechanisms. Experimentally we find that, indeed, collisional vibrational deactivation of CF3H is isotopically selective. This is, perhaps, the first direct observation of isotopically selective collisional relaxation of highly excited medium-sized molecules. At low pressures and increased time-delay between the lasers both suggested mechanisms contribute equally to the enhancement of isotopic selectivity (with a slight dominance of the IRMPD step). We also perform numerical simulations of vibrational energy transfer (VET) between highly excited and cold CF3H of both isotopic species. The model we employ includes V-V energy transfer due to long-range dipole-dipole interactions and V-V',T,R energy transfer due to head-on collisions between two molecules. The results reproduce well the measured isotopic selectivity in vibrational energy transfer. At the next step, we propose a model for the simulation of the laser isotope separation process. We solve the master equations including vibrational energy transfer and absorption and stimulated emission of IR photons. The main improvement of our calculations with respect to existing models is that we introduce a dependence of the absorption/emission rates on the frequency of the laser, the internal energy of the molecule and its isotopic species. We are able to reproduce our experiments numerically and thus gain information on the laser isotope separation process. In particular, we find that at high sample pressures the mechanism of isotopically selective IRMPD rates prevails over the mechanism of different collisional relaxation rates.

EPFL2005

Using reinforcement learning for optimizing the reproduction of tasks in robot programming by demonstration

Florent Günter

As robots start pervading human environments, the need for new interfaces that would simplify human-robot interaction has become more pressing. Robot Programming by Demonstration (RbD) develops intuitive ways of programming robots, taking inspiration in strategies used by humans to transmit knowledge to apprentices. The user-friendliness of RbD is meant to allow lay users with no prior knowledge in computer science, electronics or mechanics to train robots to accomplish tasks the same way as they would with a co-worker. When a trainer teaches a task to a robot, he/she shows a particular way of fulfilling the task. For a robot to be able to learn from observing the trainer, it must be able to learn what the task entails (i.e. answer the so-called "What-to-imitate?" question), by inferring the user's intentions. But most importantly, the robot must be able to adapt its own controller to fit at best the demonstration (the so-called "How-to-imitate?" question) despite different setups and embodiments. The latter is the question that interested us in this thesis. It relates to the problem of optimizing the reproduction of the task under environmental constraints. The "How-to-imitate?" question is subdivided into two problems. The first problem, also known as the "correspondence problem", relates to resolving the discrepancy between the human demonstrator and robot's body that prevent the robot from doing an identical reproduction of the task. Even though we helped ourselves by considering solely humanoid platforms, that is platforms that have a joint configuration similar to that of the human, discrepancies in the number of degrees of freedom and range of motion remained. We resolved these by exploiting the redundant information conveyed through the demonstrations by collecting data through different frames of reference. By exploiting these redundancies in an algorithm comparable to the damped least square algorithm, we are able to reproduce a trajectory that minimizes the error between the desired trajectory and the reproduced trajectory across each frame of reference. The second problem consists in reproducing a trajectory in an unknown setup while respecting the task constraints learned during training. When the information learned from the demonstration no longer suffice to generalize the task constraints to a new set-up, the robot must re-learn the task; this time through trial-and-error. Here we considered the combination of trial-and-error learning to complement RbD. By adding a trial-and-error module to the original Imitation Learning algorithm, the robot can find a solution that is more adapted to the context and to its embodiment than the solution found using RbD. Specifically, we compared Reinforcement Learning (RL) – to other classical optimization techniques. We show that the system is advantageous in that: a) learning is more robust to unexpected events that have not been encountered during the demonstrations and b) the robot is able to optimize its own model of the task according to its own embodiment.

EPFL2009

Evolution of longevity improves immunity in Drosophila

Bruno Lemaitre

Much has been learned about the genetics of aging from studies in model organisms, but still little is known about naturally occurring alleles that contribute to variation in longevity. For example, analysis of mutants and transgenes has identified insulin signaling as a major regulator of longevity, yet whether standing variation in this pathway underlies microevolutionary changes in lifespan and correlated fitness traits remains largely unclear. Here, we have analyzed the genomes of a set of Drosophila melanogaster lines that have been maintained under direct selection for postponed reproduction and indirect selection for longevity, relative to unselected control lines, for over 35 years. We identified many candidate loci shaped by selection for longevity and late-life fertility, but - contrary to expectation - we did not find overrepresentation of canonical longevity genes. Instead, we found an enrichment of immunity genes, particularly in the Toll pathway, suggesting that evolutionary changes in immune function might underpin - in part - the evolution of late-life fertility and longevity. To test whether this genomic signature is causative, we performed functional experiments. In contrast to control flies, long-lived flies tended to downregulate the expression of antimicrobial peptides upon infection with age yet survived fungal, bacterial, and viral infections significantly better, consistent with alleviated immunosenescence. To examine whether genes of the Toll pathway directly affect longevity, we employed conditional knockdown using in vivo RNAi. In adults, RNAi against the Toll receptor extended lifespan, whereas silencing the pathway antagonist cactus--causing immune hyperactivation - dramatically shortened lifespan. Together, our results suggest that genetic changes in the age-dependent regulation of immune homeostasis might contribute to the evolution of longer life.

JOHN WILEY & SONS LTD2018