Search

Solving high-dimensional optimal control problems in real-time is an important but challenging problem, with applications to multi-agent path planning problems, which have drawn increased attention given the growing popularity of drones in recent years. In this paper, we propose a novel neural network method called SympOCnet that applies the Symplectic network to solve high-dimensional optimal control problems with state constraints. We present several numerical results on path planning problems in two-dimensional and three-dimensional spaces. Specifically, we demonstrate that our SympOCnet can solve a problem with more than 500 dimensions in 1.5 hours on a single GPU, which shows the effectiveness and efficiency of SympOCnet. The proposed method is scalable and has the potential to solve truly high-dimensional path planning problems in real-time.

Purpose: To investigate the clinical relevance of the recently characterized human oncoprotein cancerous inhibitor of protein phosphatase 2A (CIP2A) in human breast cancer.Experimental Design: CIP2A expression (mRNA and protein) was measured in three different sets of human mammary tumors and compared with clinicopathologic variables. The functional role of CIP2A in breast cancer cells was evaluated by small interfering RNA–mediated depletion of the protein followed by an analysis of cell proliferation, migration, anchorage-independent growth, and xenograft growth.Results: CIP2A mRNA is overexpressed (n = 159) and correlates with higher Scarff-Bloom-Richardson grades (n = 251) in samples from two independent human breast cancer patients. CIP2A protein was found to be overexpressed in 39% of 33 human breast cancer samples. Furthermore, CIP2A mRNA expression positively correlated with lymph node positivity of the patients and with the expression of proliferation markers and p53 mutations in the tumor samples. Moreover, CIP2A protein expression was induced in breast cancer mouse models presenting mammary gland–specific depletion of p53 and either BRCA1 or BRCA2. Functionally, CIP2A depletion was shown to inhibit the expression of its target protein c-Myc. Loss of CIP2A also inhibited anchorage-independent growth in breast cancer cells. Finally, CIP2A was shown to support MDA-MB-231 xenograft growth in nude mice.Conclusions: Our data show that CIP2A is associated with clinical aggressivity in human breast cancer and promotes the malignant growth of breast cancer cells. Thus, these results validate the role of CIP2A as a clinically relevant human oncoprotein and warrant further investigation of CIP2A as a therapeutic target in breast cancer treatment. (Clin Cancer Res 2009;15(16):5092–100)

Purpose: The identification of a molecular signature predicting the relapse of tamoxifen-treated primary breast cancers should help the therapeutic management of estrogen receptor–positive cancers.Experimental Design: A series of 132 primary tumors from patients who received adjuvant tamoxifen were analyzed for expression profiles at the whole-genome level by 70-mer oligonucleotide microarrays. A supervised analysis was done to identify an expression signature.Results: We defined a 36-gene signature that correctly classified 78% of patients with relapse and 80% of relapse-free patients (79% accuracy). Using 23 independent tumors, we confirmed the accuracy of the signature (78%) whose relevance was further shown by using published microarray data from 60 tamoxifen-treated patients (63% accuracy). Univariate analysis using the validation set of 83 tumors showed that the 36-gene classifier is more efficient in predicting disease-free survival than the traditional histopathologic prognostic factors and is as effective as the Nottingham Prognostic Index or the “Adjuvant!” software. Multivariate analysis showed that the molecular signature is the only independent prognostic factor. A comparison with several already published signatures demonstrated that the 36-gene signature is among the best to classify tumors from both training and validation sets. Kaplan-Meier analyses emphasized its prognostic power both on the whole cohort of patients and on a subgroup with an intermediate risk of recurrence as defined by the St. Gallen criteria.Conclusion: This study identifies a molecular signature specifying a subgroup of patients who do not gain benefits from tamoxifen treatment. These patients may therefore be eligible for alternative endocrine therapies and/or chemotherapy.

Supplementary Data from CIP2A Is Associated with Human Breast Cancer Aggressivity

Supplementary Data from A Gene Expression Signature that Can Predict the Recurrence of Tamoxifen-Treated Primary Breast Cancer

Hamilton-Jacobi partial differential equations (HJ PDEs) have deep connections with a wide range of fields, including optimal control, differential games, and imaging sciences. By considering the time variable to be a higher dimensional quantity, HJ PDEs can be extended to the multi-time case. In this paper, we establish a novel theoretical connection between specific optimization problems arising in machine learning and the multi-time Hopf formula, which corresponds to a representation of the solution to certain multi-time HJ PDEs. Through this connection, we increase the interpretability of the training process of certain machine learning applications by showing that when we solve these learning problems, we also solve a multi-time HJ PDE and, by extension, its corresponding optimal control problem. As a first exploration of this connection, we develop the relation between the regularized linear regression problem and the Linear Quadratic Regulator (LQR). We then leverage our theoretical connection to adapt standard LQR solvers (namely, those based on the Riccati ordinary differential equations) to design new training approaches for machine learning. Finally, we provide some numerical examples that demonstrate the versatility and possible computational advantages of our Riccati-based approach in the context of continual learning, post-training calibration, transfer learning, and sparse dynamics identification., 26 pages in total, 23 pages for the main text, 9 figures

The hypothalamic neuropeptide oxytocin (OT) exerts prominent analgesic effects via central and peripheral action. However, the precise analgesic pathways recruited by OT are largely elusive. Here we discovered a subset of OT neurons whose projections preferentially terminate on OT receptor (OTR)-expressing neurons in the ventrolateral periaqueductal gray (vlPAG). Using a newly generated line of transgenic rats (OTR-IRES-Cre), we determined that most of the vlPAG OTR expressing cells targeted by OT projections are GABAergic. Ex vivo stimulation of parvocellular OT axons in the vlPAG induced local OT release, as measured with OT sensor GRAB. In vivo, optogenetically-evoked axonal OT release in the vlPAG of as well as chemogenetic activation of OTR vlPAG neurons resulted in a long-lasting increase of vlPAG neuronal activity. This lead to an indirect suppression of sensory neuron activity in the spinal cord and strong analgesia in both female and male rats. Altogether, we describe an OT-vlPAG-spinal cord circuit that is critical for analgesia in both inflammatory and neuropathic pain models.

Deep brain stimulation (DBS) therapies have shown clinical success in the treatment of a number of neurological illnesses, including obsessive-compulsive disorder, epilepsy, and Parkinson's disease. An emerging strategy for increasing the efficacy of DBS therapies is to develop closed-loop, adaptive DBS systems that can sense biomarkers associated with particular symptoms and in response, adjust DBS parameters in real-time. The development of such systems requires extensive analysis of the underlying neural signals while DBS is on, so that candidate biomarkers can be identified and the effects of varying the DBS parameters can be better understood. However, DBS creates high amplitude, high frequency stimulation artifacts that prevent the underlying neural signals and thus the biological mechanisms underlying DBS from being analyzed. Additionally, DBS devices often require low sampling rates, which alias the artifact frequency, and rely on wireless data transmission methods that can create signal recordings with missing data of unknown length. Thus, traditional artifact removal methods cannot be applied to this setting. We present a novel periodic artifact removal algorithm for DBS applications that can accurately remove stimulation artifacts in the presence of missing data and in some cases where the stimulation frequency exceeds the Nyquist frequency. The numerical examples suggest that, if implemented on dedicated hardware, this algorithm has the potential to be used in embedded closed-loop DBS therapies to remove DBS stimulation artifacts and hence, to aid in the discovery of candidate biomarkers in real-time. Code for our proposed algorithm is publicly available on Github.

In most

Variational and Bayesian methods are two approaches that have been widely used to solve image reconstruction problems. In this paper, we propose original connections between Hamilton--Jacobi (HJ) partial differential equations and a broad class of Bayesian methods and posterior mean estimators with Gaussian data fidelity term and log-concave prior. Whereas solutions to certain first-order HJ PDEs with initial data describe maximum a posteriori estimators in a Bayesian setting, here we show that solutions to some viscous HJ PDEs with initial data describe a broad class of posterior mean estimators. These connections allow us to establish several representation formulas and optimal bounds involving the posterior mean estimate. In particular, we use these connections to HJ PDEs to show that some Bayesian posterior mean estimators can be expressed as proximal mappings of twice continuously differentiable functions, and furthermore we derive a representation formula for these functions., Comment: 33 pages, 2 figures

Introduction: France has been impacted by the COVID-19 pandemic. Anxiety, depression, burn out and the high proportion of post-traumatic stress disorder proved to be the most expected troubles caused by this pandemic and the confinement. Medico-psychological emergency units (CUMP) have been solicited at the very early stage of the pandemic because CUMP units are very well known by the French government and systematically associated to emergency plans. Methods: In this article we describe the process which has been developed to cope with the psychological needs in the general population. At a first level, platforms of volunteers specialised into listening were available. Then those platforms could directly mobilise the CUMP in case of psychiatric disorders. It ran over the whole first wave and it has been reactivated because of the second confinement in France. Results: During the first wave, approximately 1% of all the calls made on the national Covid number required to be redirected to the listening platforms. Of this group, 4% were related to reactive pathology or a psychiatric decompensating that required adapted and specialised care. Conclusion: The high rates of psychological distress detected in the general population in recent scientific literature seem discrepant with our findings of relatively low reorientation towards the CUMP. Nevertheless, our study highlights that the response of the CUMP network in France during the first wave was supportive. The second wave displays its adaptability to the public health policies.

Deep brain stimulation (DBS) therapies have shown clinical success in the treatment of a number of neurological illnesses, including obsessive-compulsive disorder, depression, and Parkinson’s disease. An emerging strategy for increasing the efficacy of DBS therapies is to develop closed-loop, adaptive DBS systems that can sense biomarkers associated with particular symptoms and in response, adjust DBS parameters in realtime. The development of such systems requires extensive analysis of the underlying neural signals while DBS is on, so that candidate biomarkers can be identified and the effects of varying the DBS parameters can be better understood. However, DBS creates high amplitude, high frequency stimulation artifacts that prevent the underlying neural signals and thus the biological mechanisms underlying DBS from being analyzed. Additionally, DBS devices often require low sampling rates, which alias the artifact frequency, and rely on wireless data transmission methods that can create signal recordings with missing data of unknown length. Thus, traditional artifact removal methods cannot be applied to this setting. We present a novel periodic artifact removal algorithm for DBS applications that can accurately remove stimulation artifacts in the presence of missing data and in some cases where the stimulation frequency exceeds the Nyquist frequency. The numerical examples suggest that, if implemented on dedicated hardware, this algorithm has the potential to be used in embedded closed-loop DBS therapies to remove DBS stimulation artifacts and hence, to aid in the discovery of candidate biomarkers in real-time. Code for our proposed algorithm is publicly available on Github.

Este articulo analiza la composicion de la clase media brasilena, sus principales comportamientos y sus expectativas, combinando un analisis cuantitativo basado en la Encuesta Nacional de Hogares (PNAD) y una encuesta cualitativa realizada a hogares de clase media. En primer lugar, la clase media brasilena es heterogenea, y una parte sustancial de ella continua siendo vulnerable a la pobreza. En segundo lugar, el comportamiento de los consumidores de clase media es sostenido en gran medida por el credito. En tercer lugar, en terminos de expectativas, la clase media prioriza la salud, la educacion, la seguridad y la vivienda, y es particularmente critica con la calidad de los servicios publicos, tanto mas cuanto que soporta una elevada carga fiscal. Mientras que la clase media alta es capaz de soslayar estas deficiencias recurriendo a los servicios privados, los grupos mas vulnerables continuan siendo muy dependientes de los servicios publicos de baja calidad.

Physics informed neural networks (PINNs) are deep learning based techniques for solving partial differential equations (PDEs) encounted in computational science and engineering. Guided by data and physical laws, PINNs find a neural network that approximates the solution to a system of PDEs. Such a neural network is obtained by minimizing a loss function in which any prior knowledge of PDEs and data are encoded. Despite its remarkable empirical success in one, two or three dimensional problems, there is little theoretical justification for PINNs. As the number of data grows, PINNs generate a sequence of minimizers which correspond to a sequence of neural networks. We want to answer the question: Does the sequence of minimizers converge to the solution to the PDE? We consider two classes of PDEs: linear second-order elliptic and parabolic. By adapting the Schauder approach and the maximum principle, we show that the sequence of minimizers strongly converges to the PDE solution in $C^0$. Furthermore, we show that if each minimizer satisfies the initial/boundary conditions, the convergence mode becomes $H^1$. Computational examples are provided to illustrate our theoretical findings. To the best of our knowledge, this is the first theoretical work that shows the consistency of PINNs.

Actinobacteria of the genus

The hypothalamic neuropeptide, oxytocin (OT), exerts prominent analgesic effects via central and peripheral action. Here we discovered a novel subset of OT neurons whose projections preferentially terminate on OT receptor (OTR)-expressing neurons in the ventrolateral periaqueductal gray (vlPAG). Using a newly generated line of transgenic rats (OTR-IRES-Cre), we determined that most of the vlPAG OTR expressing cells being targeted by OT projections are GABAergic in nature. Both optogenetically-evoked axonal OT release in the vlPAG as well as chemogenetic activation of OTR vlPAG neurons results in a long-lasting overall increase of vlPAG neuronal activity. This then leads to an indirect suppression of sensory neuron activity in the spinal cord and strong analgesia. Finally, we describe a novel OT→vlPAG→spinal cord circuit that seems critical for analgesia in the context of both inflammatory and neuropathic pain.Highlights- We generated a new transgenic knock-in rat line (OTR-IRES-Cre)- A distinct parvOT neuronal population projects to vlPAG but not the SON or spinal cord- OT excites vlPAG OTR neurons which indirectly inhibit SC WDR neurons- This novel parvOT→vlPAG→SC pathway alleviates nociception but not the affective component of pain

To implement the pharmaceutical record in retrocession, to evaluate its contribution to the analysis of drug interactions and to estimate the sustainability of this approach.This prospective, descriptive, monocentric study was conducted over five months. All patients presenting at the retrocession were eligible. After having offered them the pharmaceutical record and having completed it, drug interactions were sought. If the impact was considered significant, a pharmaceutical intervention was transmitted to the referring physician of the institution and recorded in the computerized patient record.The pharmaceutical record was offered to 497 patients, i.e., 87 % of eligible patients. At the first meeting, 7 % of patients (n=34) were aware of it and 72 % had one open. In total, 395 pharmaceutical records were filled in at least once, 41 of which we created. Only 25 patients (5 %) refused the process and 90 % of the existing records were filled by the pharmacy. In total, 419 prescriptions were analysed for 330 patients: the pharmaceutical record was therefore a useful tool for 66 % of patients. For 17 % (n=57) of them, or 11 % of included patients, 99 drug interactions with a high risk of clinical impact were detected with the retroceded drug. On average, the presentation, creation and feeding of the drug record took one minute each and the analysis of interactions 14minutes.Easy to implement, the pharmaceutical record is a useful tool to search for drug interactions with retroceded drugs. It helps to optimize patient follow-up, despite the limited information available.

Problem:Both water and nitrogen stress in crops is a problem for agriculture across Europe and beyond. If irrigation cannot be scheduled, dry seasons lead to water stress in potato crops, resulting in yield losses and lower tuber quality. Furthermore, nitrogen stress can also cause similar problems on yields and lower tuber quality for potato crops. Nitrogen stress and deficiency in crops can be caused by excessive irrigation and heavy rains, because of nutrient leaching. The latter can also potentially contaminate local and ground water. Solution:Pre-crops are used to improve soil structure and fertilization for the subsequent crops planted in the same field area. Using adequate pre-crops require less water and nitrogen fertilizer input. In west Switzerland, rye and soybean pre-crops for following potato crops have been tested, in very sandy, well-draining soils, where water use efficiency can be low. It was observed that yield losses were minimized with the pre-crops planted before. With rye as a pre-crop, water stress lessened the yield loss from 25% in comparison with 36% after soybean. However, yield loss after nitrogen stress amounted to 4.8% after soybean and 12% after rye. These results, together with weather and soil conditions, guide the choice of a suitable pre-crop for enhanced yield performance of potato.

Many hypotheses could explain the mortality decrease observed using hemodiafiltration, such as reduction of intradialytic hypotension and more efficient toxin removal. We led a systematic analysis of representative uremic toxin removal with hemodialysis (HD), online postdilution hemodiafiltration (postHDF) and online predilution hemodiafiltration (preHDF), in a single-center crossover and prospective observational study. The primary outcome was the reduction ratio of uremic toxins of the three categories defined by the Eutox group. Twenty-six patients were treated by those three techniques of extra renal epuration. Mean Kt/Vurea was not different between the treatment methods. Mean reduction ratio of beta2microglobulin was significantly higher for both HDF treatments than for HD (p < 0.001). Myoglobin, kappa, and lambda free light chain reduction ratio was significantly different between the modes: 37.75 ± 11.95%, 45.31 ± 11% and 61.22 ± 10.56%/57.21 ± 12.5%, 63.53 ± 7.93%, and 68.40 ± 11.79%/29.12 ± 8.44%, 34.73 ± 9.01%, and 45.55 ± 12.31% HD, preHDF, and postHDF, respectively (p < 0.001). Mean protein-bound solutes reduction ratio was not different between the different treatments except for PCS with a higher reduction ratio during HDF treatments. Mean albumin loss was always less than 2 g. HDF improved removal of middle molecules but had no effect on indoles concentration without any difference between synthetic dialysis membranes.

Logistic regression is a widely used statistical model to describe the relationship between a binary response variable and predictor variables in data sets. It is often used in machine learning to identify important predictor variables. This task, variable selection, typically amounts to fitting a logistic regression model regularized by a convex combination of $\ell_1$ and $\ell_{2}^{2}$ penalties. Since modern big data sets can contain hundreds of thousands to billions of predictor variables, variable selection methods depend on efficient and robust optimization algorithms to perform well. State-of-the-art algorithms for variable selection, however, were not traditionally designed to handle big data sets; they either scale poorly in size or are prone to produce unreliable numerical results. It therefore remains challenging to perform variable selection on big data sets without access to adequate and costly computational resources. In this paper, we propose a nonlinear primal-dual algorithm that addresses these shortcomings. Specifically, we propose an iterative algorithm that provably computes a solution to a logistic regression problem regularized by an elastic net penalty in $O(T(m,n)\log(1/\epsilon))$ operations, where $\epsilon \in (0,1)$ denotes the tolerance and $T(m,n)$ denotes the number of arithmetic operations required to perform matrix-vector multiplication on a data set with $m$ samples each comprising $n$ features. This result improves on the known complexity bound of $O(\min(m^2n,mn^2)\log(1/\epsilon))$ for first-order optimization methods such as the classic primal-dual hybrid gradient or forward-backward splitting methods., Comment: 15 pages

International audience; Évoquer la route sonore de l’esclave du Bénin aux Antilles revient à passer en revue l’histoire de la traite négrière transatlantique en Afrique et au Bénin avant de restreindre le champ d’analyse aux expressions musicales des esclaves du Danxomè (ancien nom du Dahomey, autrement dit du Bénin actuel) dans les plantations de canne à sucre, de café et de coton aux Amériques et dans la Caraïbe, plus particulièrement aux Antilles françaises. De quelles expressions sonores s’agit-il? Qui sont les auteurs de ces musiques? Quelle est leur histoire? Quelles traces reste-t-il des influences de ces musiques sur les arts traditionnels des Antilles et sur les musiques actuelles de la diaspora béninoise? Une telle analysepermettra également de mieux cerner l’apport de Béninois à la musique antillaise. Sans prétendre à l’exhaustivité, cet article est un panorama exploratoire de travaux à réaliser pour motiver et développer des recherches relatives à la Route sonore de l’Esclave, depuis le port de Ouidah et sa région jusqu’aux Antilles et au-delà.

Context: Menthol, the main monoterpene found in Mentha piperita L. (M. piperita) is known to modulate nociceptive threshold and is present in different curative preparations that reduce sensory hypersensitivities in pain conditions. While for pulegone, a menthol-like monoterpene, only a limited number of studies focus on its putative analgesic effects, pulegone is the most abundant monoterpene present in Calamintha nepeta (L.) Savi (C. nepeta), a plant of the Lamiaceae family used in traditional medicine to alleviate rheumatic disorders, which counts amongst chronic inflammatory diseases.Objectives: Here, we analyzed the monoterpenes composition of C. nepeta and M. piperita. We then compared the putative anti-hyperalgesic effects of the main monoterpenes found, menthol and pulegone, in acute inflammatory pain conditions.Methods:C. nepeta and M. piperita extracts were obtained through pressurized liquid extraction and analyzed by gas chromatography-mass spectrometry. The in vitro anti-inflammatory activity of menthol or pulegone was evaluated by measuring the secretion of the tumour necrosis factor alpha (TNF α) from LPS-stimulated THP-1 cells. The in vivo anti-hyperalgesic effects of menthol and pulegone were tested on a rat inflammatory pain model.Results: Pulegone and menthol are the most abundant monoterpene found in C. nepeta (49.41%) and M. piperita (42.85%) extracts, respectively. In vitro, both pulegone and menthol act as strong anti-inflammatory molecules, with EC50 values of 1.2 ± 0.2 and 1.5 ± 0.1 mM, respectively, and exert cytotoxicity with EC50 values of 6.6 ± 0.3 and 3.5 ± 0.2 mM, respectively. In vivo, 100 mg/kg pulegone exerts a transient anti-hyperalgesic effect on both mechanical (pulegone: 274.25 ± 68.89 g, n = 8; vehicle: 160.88 ± 35.17 g, n = 8, p < 0.0001), thermal heat (pulegone: 4.09 ± 0.62 s, n = 8; vehicle: 2.25 ± 0.34 s, n = 8, p < 0.0001), and cold (pulegone: 2.25 ± 1.28 score, n = 8; vehicle: 4.75 ± 1.04 score, n = 8, p = 0.0003). In a similar way, 100 mg/kg menthol exerts a transient anti-hyperalgesic effect on both mechanical (mechanical: menthol: 281.63 ± 45.52 g, n = 8; vehicle: 166.25 ± 35.4 g, n = 8, p < 0.0001) and thermal heat (menthol: 3.65 ± 0.88 s, n = 8; vehicle: 2.19 ± 0.26 s, n = 8, Conclusion: Here, we show that both pulegone and menthol are anti-inflammatory and anti-hyperalgesic monoterpenes. These results might open the path towards new compound mixes to alleviate the pain sensation.

We report the development of a high-brightness, high-power Ce:YAG luminescent concentrator pumped by 2240 blue LEDs in quasi-continuous wave operation (10 µs, 10 Hz). Using light confinement and recycling in the three space dimensions, the parallelepiped (1mm×14×mm×200mm) Ce:YAG emits a power of 145 W from a square output surface (1 × 1mm2) corresponding to a brightness of 4.6 kW/cm2/sr. This broadband yellow source has a unique combination of luminous flux (7.6 104 lm) and brightness (2.4 104 cd/mm2) and overcomes many other visible incoherent sources by one order of magnitude. This paper also proposes a deep understanding of the performance drop compared to a linear behavior when the pump power increases. Despite excited state absorption was unexpected for this low doped Ce:YAG pumped at a low irradiance level, we demonstrated that it affects the performance by tripling the losses in the concentrator. This effect is particularly important for small output surfaces corresponding to strong light recycling in the concentrator and to average travel distances inside the medium reaching meters.

Two key challenges in optimal control include efficiently solving high-dimensional problems and handling optimal control problems with state-dependent running costs. In this paper, we consider a class of optimal control problems whose running costs consist of a quadratic on the control variable and a convex, non-negative, piecewise affine function on the state variable. We provide the analytical solution for this class of optimal control problems as well as a Hopf-type representation formula for the corresponding Hamilton-Jacobi partial differential equations. Finally, we propose efficient numerical algorithms based on our Hopf-type representation formula, convex optimization algorithms, and min-plus techniques. We present several high-dimensional numerical examples, which demonstrate that our algorithms overcome the curse of dimensionality. We also describe a field-programmable gate array (FPGA) implementation of our numerical solver whose latency scales linearly in the spatial dimension and that achieves approximately a 40 times speedup compared to a parallelized central processing unit (CPU) implementation. Thus, our numerical results demonstrate the promising performance boosts that FPGAs are able to achieve over CPUs. As such, our proposed methods have the potential to serve as a building block for solving more complicated high-dimensional optimal control problems in real-time., arXiv admin note: text overlap with arXiv:2109.14849

The neonatal period is critical for brain development and determinant for long-term brain trajectory. Yet, this time concurs with a sensitivity and risk for numerous brain injuries following perinatal complications such as preterm birth. Brain injury in premature infants leads to a complex amalgam of primary destructive diseases and secondary maturational and trophic disturbances and, as a consequence, to long-term neurocognitive and behavioral problems. Neuroinflammation is an important common factor in these complications, which contributes to the adverse effects on brain development. Mediating this inflammatory response forms a key therapeutic target in protecting the vulnerable developing brain when complications arise. The neuropeptide oxytocin (OT) plays an important role in the perinatal period, and its importance for lactation and social bonding in early life are well-recognized. Yet, novel functions of OT for the developing brain are increasingly emerging. In particular, OT seems able to modulate glial activity in neuroinflammatory states, but the exact mechanisms underlying this connection are largely unknown. The current review provides an overview of the oxytocinergic system and its early life development across rodent and human. Moreover, we cover the most up-to-date understanding of the role of OT in neonatal brain development and the potential neuroprotective effects it holds when adverse neural events arise in association with neuroinflammation. A detailed assessment of the underlying mechanisms between OT treatment and astrocyte and microglia reactivity is given, as well as a focus on the amygdala, a brain region of crucial importance for socio-emotional behavior, particularly in infants born preterm.

We detail results of an experiment performed at the Ligne d'Intégration Laser facility aimed at studying supersonic and diffusive radiation front propagation in low-density SiO2 aerogel (20 and 40 mg/cm3) enclosed in a gold tube, driven by thermal emission from a laser-heated spherical gold cavity. The evolution of the front is studied continuously by measuring its self-emission with a 1D (one-dimensional) time-resolved soft x-ray imager. Measurement is performed along (through a 200- μm-wide observation slit) and at the exit of the tube giving access to the dynamics and the curvature of the front. Experimental results are then compared successfully to results from the 3D (three-dimensional) radiation hydrodynamics code TROLL, which shows that if continuous tracking of the front position is accessible with this experimental scheme, measurement of its maximum radiation temperature is on the contrary affected by radiation closure of the observation slit. 3D simulations also indicate that this effect can even be worsened if one includes pointing errors of the x-ray imager. Radiation temperature along the tube was then inferred by combining results from the imager to a wall shock breakout time measurement using a velocity interferometer system for any reflector and results from a broadband x-ray spectrometer used to determine the temperature at the exit of the tube. A decrease in the radiation temperature along the tube is observed, the decrease being more important for the higher SiO2 aerogel density.

International audience; The term “middle class” is increasingly used to qualify the mass of people in developing countries who are neither poor nor rich and share consumption patterns historically associated with the western middle class. However, what differentiates them from the ideal-typical middle class, as well as the extent to which emerging middle classes differ across developing countries, has only been marginally documented by economists to date. This article proposes to scratch beneath the surface of the so-called middle-class that is burgeoning in developing countries by documenting the commonalities and differences hidden by the all-encompassing term of “middle class” for a set of developing countries exhibiting different levels of income, economic structures and socio-political systems: Brazil, Côte d’Ivoire, Turkey and Vietnam. Relying on quantitative and qualitative micro-economic data, our paper compares the objective characteristics (occupation, education, income), behavior and subjective expectations of the people standing in the middle of the income distribution in the four countries. Four main research questions structure the paper. Where is the middle? How is the middle specific? Who is in the middle? What does the middle aspire to? Results show that the middle-income group in each country covers distinct realities. Although some similar characteristics and behaviors are observable, our results reveal a strong heterogeneity within each national middle class, with no fewer than four to seven socio-economic subgroups, and strong country-specific traits as most of the subgroups are deeply rooted in the country’s specific historical trajectory. In each country, our analysis also unveils a pattern of bipolarization between a rather affluent and stable middle class and a “new” or more vulnerable one. Finally, middle class members appear to be characterized by an individualist positioning and the absence of a marked political commitment, challenging the common assumption that developing countries’ middle classes are agents of political change.

Two of the main challenges in optimal control are solving problems with state-dependent running costs and developing efficient numerical solvers that are computationally tractable in high dimension. In this paper, we provide analytical solutions to certain optimal control problems whose running cost depends on the state variable and with constraints on the control. We also provide Lax-Oleinik-type representation formulas for the corresponding Hamilton-Jacobi partial differential equations with state-dependent Hamiltonians. Additionally, we present an efficient, grid-free numerical solver based on our representation formulas, which is shown to scale linearly with the state dimension, and thus, to overcome the curse of dimensionality. Using existing optimization methods and the min-plus technique, we extend our numerical solvers to address more general classes of convex and nonconvex initial costs. We demonstrate the capabilities of our numerical solvers using implementations on a central processing unit (CPU) and a field-programmable gate array (FPGA). In several cases, our FPGA implementation obtains over a 10 times speedup compared to the CPU, which demonstrates the promising performance boosts FPGAs can achieve. Our numerical results show that our solvers have the potential to serve as a building block for solving broader classes of high-dimensional optimal control problems in real-time.

The linear primal-dual hybrid gradient (PDHG) method is a first-order method that splits convex optimization problems with saddle-point structure into smaller subproblems. Unlike those obtained in most splitting methods, these subproblems can generally be solved efficiently because they involve simple operations such as matrix-vector multiplications or proximal mappings that are fast to evaluate numerically. This advantage comes at the price that the linear PDHG method requires precise stepsize parameters for the problem at hand to achieve an optimal convergence rate. Unfortunately, these stepsize parameters are often prohibitively expensive to compute for large-scale optimization problems, such as those in machine learning. This issue makes the otherwise simple linear PDHG method unsuitable for such problems, and it is also shared by most first-order optimization methods as well. To address this issue, we introduce accelerated nonlinear PDHG methods that achieve an optimal convergence rate with stepsize parameters that are simple and efficient to compute. We prove rigorous convergence results, including results for strongly convex or smooth problems posed on infinite-dimensional reflexive Banach spaces. We illustrate the efficiency of our methods on $\ell_{1}$-constrained logistic regression and entropy-regularized matrix games. Our numerical experiments show that the nonlinear PDHG methods are considerably faster than competing methods., 52 pages, no figures

Actinobacteria belonging to the genus Amycolatopsis are important for antibiotic production and other valuable biotechnological applications such as biodegradation or bioconversion. Despite their industrial importance, tools and methods for the genetic manipulation of Amycolatopsis are less developed than in other actinobacteria such as Streptomyces. Moreover, most of the existing methods do not support convenient marker-free genome engineering. Here, we report the use of the pSAM2 site-specific recombination system for the efficient deletion of marker genes or large DNA regions in Amycolatopsis. For this purpose, we constructed a shuttle vector, replicating in Escherichia coli and Amycolatopsis, expressing the Xis and Int proteins from the Streptomyces integrative and conjugative element pSAM2. These proteins are sufficient for site-specific recombination between the attachment sites attL and attR. We also constructed two plasmids, replicative in E. coli but not in Amycolatopsis, for the integration of the recombination sites attL and attR on each side of a region targeted for deletion. We exemplified the use of these tools in Amycolatopsis mediterranei DSM 40773 by obtaining with high efficiency (>95%) a marker-free deletion of one single gene in the rifamycin biosynthetic gene cluster or of the entire 90-kb cluster.IMPORTANCEThe genus Amycolatopsis is regarded as an important source of diverse specialized metabolites. Members of this genus are used in industry for the production of valuable antibiotics such as rifamycins or vancomycin. Amycolatopsis spp. also present a great interest for biotechnological applications such as biodegradation or bioconversion. Despite their importance, their genetic manipulation was somehow hampered by the lack of efficient tools. Here we report the successful use of the pSAM2 site-specific recombination system to construct unmarked deletion mutants, allowing marker recycling, or to create large deletions in A. mediterranei DSM 40773. The high efficiency of this site-specific recombination system and it possible application to other Amycolatopsis species open new opportunities for marker-free genome engineering in this genus.

Platelets have a well-established role in atherosclerosis and related diseases. Lebetins from the venom of Vipera lebetina, lacking the RGD sequence, emerged as a new family of platelet aggregation inhibitors. However, the interaction sites and precise mechanism between lebetin and its substrate remain unclear. Here, we successfully synthesized two peptide analogs, which differ only by one glycine residue at the N-terminus: lebetin 2α (sL2α residues) and lebetin 2β (sL2ββ residues) were produced in sufficient quantities for structural and functional studies. NMR structure determination showed that the sL2α peptide adopts a compact ring conformation stabilized by a disulfide bond, from which emerge one loop and two extended regions, the C- and N-termini. Interestingly, two RGD-like motifs were identified in the structure of the peptides, suggesting an anti-platelet aggregation effect of the two isoforms. Indeed, activity was demonstrated on human and rabbit platelet-rich plasma where sL2α and sL2β showed more potent inhibitory effect on platelet aggregation compared to the previously described native lebetin 1. Synthetic lebetin 2 peptides constitute promising candidates for drug design toward chimeric compounds with high anti-platelet and natriuretic effects. These findings contribute to a novel field of research triggering platelet activation and natriuretic action.

International audience; In most Streptomyces species, antibiotic production is triggered in a condition of phosphate limitation, a condition that is known to be correlated with a low intracellular ATP content compared to growth in a condition of phosphate proficiency. This observation suggests that a low ATP content might be a direct trigger of antibiotic biosynthesis. In order to test this hypothesis, we introduced into the model strain Streptomyces lividans, a functional and a non-functional ATPase cloned into the replicative vector pOSV206 and expressed under the control of the strong ErmE* promoter. The functional ATPase was constituted by the α (AtpA), β (AtpB) and γ (AtpD) sub-units of the native F1 part of the ATP synthase of S. lividans that, when separated from the membrane-bound F0 part, bears an ATPase activity. The non-functional ATPase was a mutated version of the latter, bearing a 12 amino acids deletion encompassing the active site of the AtpD sub-unit. S. lividans was chosen to test our hypothesis since this strain hardly produces any antibiotics. However, it possesses the same biosynthetic pathways of various specialized metabolites as S. coelicolor, a phylogenetically closely related strain that produces these metabolites in abundance. Our results demonstrated that the over-expression of the functional ATPase, but not that of its mutated version, indeed correlated with the production of the bioactive metabolites of the CDA, RED and ACT clusters. These results confirmed the long known and mysterious link existing between a phosphate limitation leading to an ATP deficit and the triggering of antibiotic biosynthesis. Based on this work and the previous published results of our group, we propose an entirely novel conception of the nature of this link

Monoterpenes are small molecules, composed of two isoprene units, able to pass through the blood brain barrier, allowing to target both peripheral and central pain pathways. They are the main components of essential oils, responsible for their diverse well-known biological activities. Menthol, the main monoterpene found in Mentha piperita (L.) is known to modulate nociceptive threshold and is present in different curative preparations that reduces sensory hypersensitivities in pain conditions. While pulegone is a menthol-like monoterpene, only a limited number of studies focuses on its putative analgesic effects. Pulegone is the most abundant monoterpene presents in Calamintha nepeta (L.), a Laminaceae plant used in traditional medicine to alleviate rheumatic disorders, a chronic inflammatory disease. Here, we compared the impacts of menthol and pulegone on pain and inflammation. First, we described that both monoterpenes are anti-inflammatory compounds. Secondly, we found that while menthol is highly cytotoxic at anti-inflammatory concentrations, the cytotoxic effects of pulegone are limited, if not absent. Finally, in a model of peripheral inflammatory-induced pain a pulegone treatment exerts a significantly higher anti-hyperalgesic effect than menthol in response to mechanical stimuli, heat and cold thermal stimulations than a menthol treatment. In conclusion we demonstrated that pulegone is an anti-inflammatory compound and it is acting a potent pain-killer in acute inflammatory pain condition.

À Jean-Luc Tamby Lumumba : Il y a dans ma poitrine un dur noyau, le silex contre quoi s’ébréchera ta lame ! C’est l’honneur de l’Afrique ! […] Le banquier : Pour ma part, je ne vois là que matière à spéculation politique. C’est un épisode disons folklorique. Aimé Césaire Sera le temps Ix Jaguar où tombera l’ordre des unités du katun, l’heure de défaire le pagne, d’ôter les vêtements. […] Chilam Balam de Chamayel En 1954, Charles Ives mourut en laissant sa Symphonie de l’Univers inachev...

We consider image denoising problems formulated as variational problems. It is known that Hamilton-Jacobi PDEs govern the solution of such optimization problems when the noise model is additive. In this work, we address certain non-additive noise models and show that they are also related to Hamilton-Jacobi PDEs. These findings allow us to establish new connections between additive and non-additive noise imaging models. Specifically, we study how the solutions to these optimization problems depend on the parameters and the observed images. We show that the optimal values are ruled by some Hamilton-Jacobi PDEs, while the optimizers are characterized by the spatial gradient of the solution to the Hamilton-Jacobi PDEs. Moreover, we use these relations to investigate the asymptotic behavior of the variational model as the parameter goes to infinity, that is, when the influence of the noise vanishes. With these connections, some non-convex models for non-additive noise can be solved by applying convex optimization algorithms to the equivalent convex models for additive noise. Several numerical results are provided for denoising problems with Poisson noise or multiplicative noise.

Many imaging problems can be formulated as inverse problems expressed as finite-dimensional optimization problems. These optimization problems generally consist of minimizing the sum of a data fidelity and regularization terms. In [23,26], connections between these optimization problems and (multi-time) Hamilton--Jacobi partial differential equations have been proposed under the convexity assumptions of both the data fidelity and regularization terms. In particular, under these convexity assumptions, some representation formulas for a minimizer can be obtained. From a Bayesian perspective, such a minimizer can be seen as a maximum a posteriori estimator. In this chapter, we consider a certain class of non-convex regularizations and show that similar representation formulas for the minimizer can also be obtained. This is achieved by leveraging min-plus algebra techniques that have been originally developed for solving certain Hamilton--Jacobi partial differential equations arising in optimal control. Note that connections between viscous Hamilton--Jacobi partial differential equations and Bayesian posterior mean estimators with Gaussian data fidelity terms and log-concave priors have been highlighted in [25]. We also present similar results for certain Bayesian posterior mean estimators with Gaussian data fidelity and certain non-log-concave priors using an analogue of min-plus algebra techniques.

We propose a novel Caputo fractional derivative-based optimization algorithm. Upon defining the Caputo fractional gradient with respect to the Cartesian coordinate, we present a generic Caputo fractional gradient descent (CFGD) method. We prove that the CFGD yields the steepest descent direction of a locally smoothed objective function. The generic CFGD requires three parameters to be specified, and a choice of the parameters yields a version of CFGD. We propose three versions -- non-adaptive, adaptive terminal and adaptive order. By focusing on quadratic objective functions, we provide a convergence analysis. We prove that the non-adaptive CFGD converges to a Tikhonov regularized solution. For the two adaptive versions, we derive error bounds, which show convergence to integer-order stationary point under some conditions. We derive an explicit formula of CFGD for quadratic functions. We computationally found that the adaptive terminal (AT) CFGD mitigates the dependence on the condition number in the rate of convergence and results in significant acceleration over gradient descent (GD). For non-quadratic functions, we develop an efficient implementation of CFGD using the Gauss-Jacobi quadrature, whose computational cost is approximately proportional to the number of the quadrature points and the cost of GD. Our numerical examples show that AT-CFGD results in acceleration over GD, even when a small number of the Gauss-Jacobi quadrature points (including a single point) is used.

An amendment to this paper has been published and can be accessed via the original article.

In Streptomyces, antibiotic biosynthesis is triggered in phosphate limitation that is usually correlated with energetic stress. Polyphosphates constitute an important reservoir of phosphate and energy and a better understanding of their role in the regulation of antibiotic biosynthesis is of crucial importance. We previously characterized a gene, SLI_4384/ppk, encoding a polyphosphate kinase, whose disruption greatly enhanced the weak antibiotic production of Streptomyces lividans. In the condition of energetic stress, Ppk utilizes polyP as phosphate and energy donor, to generate ATP from ADP. In this paper, we established that ppk is co-transcribed with its two downstream genes, SLI_4383, encoding a phosin called PptA possessing a CHAD domain constituting a polyphosphate binding module and SLI_4382 encoding a nudix hydrolase. The expression of the ppk/pptA/SLI_4382 operon was shown to be under the positive control of the two-component system PhoR/PhoP and thus mainly expressed in condition of phosphate limitation. However, pptA and SLI_4382 can also be transcribed alone from their own promoter. The deletion of pptA resulted into earlier and stronger actinorhodin production and lower lipid content than the disruption of ppk, whereas the deletion of SLI_4382 had no obvious phenotypical consequences. The disruption of ppk was shown to have a polar effect on the expression of pptA, suggesting that the phenotype of the ppk mutant might be linked, at least in part, to the weak expression of pptA in this strain. Interestingly, the expression of phoR/phoP and that of the genes of the pho regulon involved in phosphate supply or saving were strongly up-regulated in pptA and ppk mutants, revealing that both mutants suffer from phosphate stress. Considering the presence of a polyphosphate binding module in PptA, but absence of similarities between PptA and known exo-polyphosphatases, we proposed that PptA constitutes an accessory factor for exopolyphosphatases or general phosphatases involved in the degradation of polyphosphates into phosphate.

In Streptomyces, antibiotic biosynthesis is triggered in phosphate limitation that is usually correlated with energetic stress. Polyphosphates constitute an important reservoir of phosphate and energy and a better understanding of their role in the regulation of antibiotic biosynthesis is of crucial importance. We previously characterized a gene, SLI_4384/ppk, encoding a polyphosphate kinase, whose disruption greatly enhanced the weak antibiotic production of Streptomyces lividans. In the condition of energetic stress, Ppk utilizes polyP as phosphate and energy donor, to generate ATP from ADP. In this paper, we established that ppk is co-transcribed with its two downstream genes, SLI_4383, encoding a phosin called PptA possessing a CHAD domain constituting a polyphosphate binding module and SLI_4382 encoding a nudix hydrolase. The expression of the ppk/pptA/SLI_4382 operon was shown to be under the positive control of the two-component system PhoR/PhoP and thus mainly expressed in condition of phosphate limitation. However, pptA and SLI_4382 can also be transcribed alone from their own promoter. The deletion of pptA resulted into earlier and stronger actinorhodin production and lower lipid content than the disruption of ppk, whereas the deletion of SLI_4382 had no obvious phenotypical consequences. The disruption of ppk was shown to have a polar effect on the expression of pptA, suggesting that the phenotype of the ppk mutant might be linked, at least in part, to the weak expression of pptA in this strain. Interestingly, the expression of phoR/phoP and that of the genes of the pho regulon involved in phosphate supply or saving were strongly up-regulated in pptA and ppk mutants, revealing that both mutants suffer from phosphate stress. Considering the presence of a polyphosphate binding module in PptA, but absence of similarities between PptA and known exo-polyphosphatases, we proposed that PptA constitutes an accessory factor for exopolyphosphatases or general phosphatases involved in the degradation of polyphosphates into phosphate.