Showing papers by "French Institute for Research in Computer Science and Automation published in 2022"

Enhancing bioreactor arrays for automated measurements and reactive control with ReacSight

Abstract: Small-scale, low-cost bioreactors provide exquisite control of environmental parameters of microbial cultures over long durations. Their use is gaining popularity in quantitative systems and synthetic biology. However, existing setups are limited in their measurement capabilities. Here, we present ReacSight, a strategy to enhance bioreactor arrays for automated measurements and reactive experiment control. ReacSight leverages low-cost pipetting robots for sample collection, handling and loading, and provides a flexible instrument control architecture. We showcase ReacSight capabilities on three applications in yeast. First, we demonstrate real-time optogenetic control of gene expression. Second, we explore the impact of nutrient scarcity on fitness and cellular stress using competition assays. Third, we perform dynamic control of the composition of a two-strain consortium. We combine custom or chi.bio reactors with automated cytometry. To further illustrate ReacSight's genericity, we use it to enhance plate-readers with pipetting capabilities and perform repeated antibiotic treatments on a bacterial clinical isolate.

kmtricks: efficient and flexible construction of Bloom filters for large sequencing data collections

Abstract: When indexing large collections of short-read sequencing data, a common operation that has now been implemented in several tools (Sequence Bloom Trees and variants, BIGSI) is to construct a collection of Bloom filters, one per sample. Each Bloom filter is used to represent a set of k-mers which approximates the desired set of all the non-erroneous k-mers present in the sample. However, this approximation is imperfect, especially in the case of metagenomics data. Erroneous but abundant k-mers are wrongly included, and non-erroneous but low-abundant ones are wrongly discarded. We propose kmtricks, a novel approach for generating Bloom filters from terabase-sized collections of sequencing data. Our main contributions are (i) an efficient method for jointly counting k-mers across multiple samples, including a streamlined Bloom filter construction by directly counting, partitioning and sorting hashes instead of k-mers, which is approximately four times faster than state-of-the-art tools; (ii) a novel technique that takes advantage of joint counting to preserve low-abundant k-mers present in several samples, improving the recovery of non-erroneous k-mers. Our experiments highlight that this technique preserves around 8× more k-mers than the usual yet crude filtering of low-abundance k-mers in a large metagenomics dataset.https://github.com/tlemane/kmtricks.Supplementary data are available at Bioinformatics Advances online.

Reliable crack detection in a rotor system with uncertainties via advanced simulation models based on kriging and Polynomial Chaos Expansion

Abstract: In this study the authors propose to take into account the nonlinear effects induced by the presence of a transverse crack to carry out vibratory monitoring and detect transverse cracks in rotating systems subject to model uncertainties. More precisely, we focus more particularly on the global complexity of the nonlinear dynamic behaviour of cracked rotors and the evolution of their harmonic components as a function of the parameters of a transverse breathing crack (its position and depth) when numerous uncertainties are considered. These random uncertainties correspond to random geometric imperfections (two disc thicknesses), random material properties (Young modulus and material density) and boundary conditions uncertainty (two bearing stiffnesses). The objective of the present work is to identify robust indicators capable of determining the presence of a crack and its status even though numerous uncertainties are present. To conduct such a study, an advanced surrogate modelling technique based on kriging and Polynomial Chaos Expansion (PCE) is proposed for the prediction of both the critical speeds and the harmonic components n × during passage through sub-critical resonances. An extensive study to ensure the validation of the surrogate models and a relevant choice of both the parametric and random Design of Experiments (i.e. kriging DoE and PCE DoE) is proposed. The proposed methodology is applied on a flexible rotor with a transverse breathing crack and subjected to random geometric imperfections and fluctuations in material properties of the rotor system.

EnosLib: A Library for Experiment-Driven Research in Distributed Computing

Abstract: Despite the importance of experiment-driven research in the distributed computing community, there has been little progress in helping researchers conduct their experiments. In most cases, they have to achieve tedious and time-consuming development and instrumentation activities to deal with the specifics of testbeds and the system under study. In order to relieve researchers of the burden of those efforts, we have developed EnosLib : a Python library that takes into account best experimentation practices and leverages modern toolkits on automatic deployment and configuration systems. EnosLib helps researchers not only in the process of developing their experimental artifacts, but also in running them over different infrastructures. To demonstrate the relevance of our library, we discuss three experimental engines built on top of EnosLib , and used to conduct empirical studies on complex software stacks between 2016 and 2019 (database systems, communication buses and OpenStack). By introducing EnosLib , our goal is to gather academic and industrial actors of our community around a library that aggregates everyday experiment-driven research operations. A library that has been already adopted by open-source projects and members of the scientific community thanks to its ease of use and extension.

From Dose Reduction to Contrast Maximization

Abstract: Objectives The aim of this study was to evaluate a deep learning method designed to increase the contrast-to-noise ratio in contrast-enhanced gradient echo T1-weighted brain magnetic resonance imaging (MRI) acquisitions. The processed images are quantitatively evaluated in terms of lesion detection performance. Materials and Methods A total of 250 multiparametric brain MRIs, acquired between November 2019 and March 2021 at Gustave Roussy Cancer Campus (Villejuif, France), were considered for inclusion in this retrospective monocentric study. Independent training (107 cases; age, 55 ± 14 years; 58 women) and test (79 cases; age, 59 ± 14 years; 41 women) samples were defined. Patients had glioma, brain metastasis, meningioma, or no enhancing lesion. Gradient echo and turbo spin echo with variable flip angles postcontrast T1 sequences were acquired in all cases. For the cases that formed the training sample, “low-dose” postcontrast gradient echo T1 images using 0.025 mmol/kg injections of contrast agent were also acquired. A deep neural network was trained to synthetically enhance the low-dose T1 acquisitions, taking standard-dose T1 MRI as reference. Once trained, the contrast enhancement network was used to process the test gradient echo T1 images. A read was then performed by 2 experienced neuroradiologists to evaluate the original and processed T1 MRI sequences in terms of contrast enhancement and lesion detection performance, taking the turbo spin echo sequences as reference. Results The processed images were superior to the original gradient echo and reference turbo spin echo T1 sequences in terms of contrast-to-noise ratio (44.5 vs 9.1 and 16.8; P < 0.001), lesion-to-brain ratio (1.66 vs 1.31 and 1.44; P < 0.001), and contrast enhancement percentage (112.4% vs 85.6% and 92.2%; P < 0.001) for cases with enhancing lesions. The overall image quality of processed T1 was preferred by both readers (graded 3.4/4 on average vs 2.7/4; P < 0.001). Finally, the proposed processing improved the average sensitivity of gradient echo T1 MRI from 88% to 96% for lesions larger than 10 mm (P = 0.008), whereas no difference was found in terms of the false detection rate (0.02 per case in both cases; P > 0.99). The same effect was observed when considering all lesions larger than 5 mm: sensitivity increased from 70% to 85% (P < 0.001), whereas false detection rates remained similar (0.04 vs 0.06 per case; P = 0.48). With all lesions included regardless of their size, sensitivities were 59% and 75% for original and processed T1 images, respectively (P < 0.001), and the corresponding false detection rates were 0.05 and 0.14 per case, respectively (P = 0.06). Conclusion The proposed deep learning method successfully amplified the beneficial effects of contrast agent injection on gradient echo T1 image quality, contrast level, and lesion detection performance. In particular, the sensitivity of the MRI sequence was improved by up to 16%, whereas the false detection rate remained similar.

Crystal or jelly? Effect of color on the perception of translucent materials with photographs of real-world objects

Abstract: Translucent materials are ubiquitous in nature (e.g. teeth, food, and wax), but our understanding of translucency perception is limited. Previous work in translucency perception has mainly used monochromatic rendered images as stimuli, which are restricted by their diversity and realism. Here, we measure translucency perception with photographs of real-world objects. Specifically, we use three behavior tasks: binary classification of "translucent" versus "opaque," semantic attribute rating of perceptual qualities (see-throughness, glossiness, softness, glow, and density), and material categorization. Two different groups of observers finish the three tasks with color or grayscale images. We find that observers' agreements depend on the physical material properties of the objects such that translucent materials generate more interobserver disagreements. Further, there are more disagreements among observers in the grayscale condition in comparison to that in the color condition. We also discover that converting images to grayscale substantially affects the distributions of attribute ratings for some images. Furthermore, ratings of see-throughness, glossiness, and glow could predict individual observers' binary classification of images in both grayscale and color conditions. Last, converting images to grayscale alters the perceived material categories for some images such that observers tend to misjudge images of food as non-food and vice versa. Our result demonstrates that color is informative about material property estimation and recognition. Meanwhile, our analysis shows that mid-level semantic estimation of material attributes might be closely related to high-level material recognition. We also discuss individual differences in our results and highlight the importance of such consideration in material perception.

Reverse Engineering Language Acquisition with Child-Centered Long-Form Recordings

Abstract: Language use in everyday life can be studied using lightweight, wearable recorders that collect long-form recordings—that is, audio (including speech) over whole days. The hardware and software und...

On the use of formal methods to model and verify neuronal archetypes

Abstract: Having a formal model of neural networks can greatly help in understanding and verifying their properties, behavior, and response to external factors such as disease and medicine. In this paper, we adopt a formal model to represent neurons, some neuronal graphs, and their composition. Some specific neuronal graphs are known for having biologically relevant structures and behaviors and we call them archetypes. These archetypes are supposed to be the basis of typical instances of neuronal information processing. In this paper we study six fundamental archetypes (simple series, series with multiple outputs, parallel composition, negative loop, inhibition of a behavior, and contralateral inhibition), and we consider two ways to couple two archetypes: (i) connecting the output(s) of the first archetype to the input(s) of the second archetype and (ii) nesting the first archetype within the second one. We report and compare two key approaches to the formal modeling and verification of the proposed neuronal archetypes and some selected couplings. The first approach exploits the synchronous programming language Lustre to encode archetypes and their couplings, and to express properties concerning their dynamic behavior. These properties are verified thanks to the use of model checkers. The second approach relies on a theorem prover, the Coq Proof Assistant, to prove dynamic properties of neurons and archetypes.

On the use of random forest for two-sample testing

Abstract: Following the line of classification-based two-sample testing, tests based on the Random Forest classifier are proposed. The developed tests are easy to use, require almost no tuning, and are applicable for any distribution on Rd. Furthermore, the built-in variable importance measure of the Random Forest gives potential insights into which variables make out the difference in distribution. An asymptotic power analysis for the proposed tests is conducted. Finally, two real-world applications illustrate the usefulness of the introduced methodology. To simplify the use of the method, the R-package “hypoRF” is provided.

Orientations and the Supersingular Endomorphism Ring Problem

Abstract: We study two important families of problems in isogeny-based cryptography and how they relate to each other: computing the endomorphism ring of supersingular elliptic curves, and inverting the action of class groups on oriented supersingular curves. We prove that these two families of problems are closely related through polynomial-time reductions, assuming the generalised Riemann hypothesis. We identify two classes of essentially equivalent problems. The first class corresponds to the problem of computing the endomorphism ring of oriented curves. The security of a large family of cryptosystems (such as CSIDH) reduces to (and sometimes from) this class, for which there are heuristic quantum algorithms running in subexponential time. The second class corresponds to computing the endomorphism ring of orientable curves. The security of essentially all isogeny-based cryptosystems reduces to (and sometimes from) this second class, for which the best known algorithms are still exponential. Some of our reductions not only generalise, but also strengthen previously known results. For instance, it was known that in the particular case of curves defined over $$\mathbf {F}_p$$ , the security of CSIDH reduces to the endomorphism ring problem in subexponential time. Our reductions imply that the security of CSIDH is actually equivalent to the endomorphism ring problem, under polynomial time reductions (circumventing arguments that proved such reductions unlikely).

Reverse engineering of a mechanistic model of gene expression using metastability and temporal dynamics

Abstract: Differentiation can be modeled at the single cell level as a stochastic process resulting from the dynamical functioning of an underlying Gene Regulatory Network (GRN), driving stem or progenitor cells to one or many differentiated cell types. Metastability seems inherent to differentiation process as a consequence of the limited number of cell types. Moreover, mRNA is known to be generally produced by bursts, which can give rise to highly variable non-Gaussian behavior, making the estimation of a GRN from transcriptional profiles challenging. In this article, we present CARDAMOM (Cell type Analysis from scRna-seq Data achieved from a Mixture MOdel), a new algorithm for inferring a GRN from timestamped scRNA-seq data, which crucially exploits these notions of metastability and transcriptional bursting. We show that such inference can be seen as the successive resolution of as many regression problem as timepoints, after a preliminary clustering of the whole set of cells with regards to their associated bursts frequency. We demonstrate the ability of CARDAMOM to infer a reliable GRN from in silico expression datasets, with good computational speed. To the best of our knowledge, this is the first description of a method which uses the concept of metastability for performing GRN inference.

Beyond Quadratic Speedups in Quantum Attacks on Symmetric Schemes

Abstract: In this paper, we report the first quantum key-recovery attack on a symmetric block cipher design, using classical queries only, with a more than quadratic time speedup compared to the best classical attack. We study the 2XOR-Cascade construction of Gaži and Tessaro (EUROCRYPT 2012). It is a key length extension technique which provides an n-bit block cipher with $$\frac{5n}{2}$$ bits of security out of an n-bit block cipher with 2n bits of key, with a security proof in the ideal model. We show that the offline-Simon algorithm of Bonnetain et al. (ASIACRYPT 2019) can be extended to, in particular, attack this construction in quantum time $$\widetilde{\mathcal {O}}\left( 2^n \right) $$ , providing a 2.5 quantum speedup over the best classical attack. Regarding post-quantum security of symmetric ciphers, it is commonly assumed that doubling the key sizes is a sufficient precaution. This is because Grover’s quantum search algorithm, and its derivatives, can only reach a quadratic speedup at most. Our attack shows that the structure of some symmetric constructions can be exploited to overcome this limit. In particular, the 2XOR-Cascade cannot be used to generically strengthen block ciphers against quantum adversaries, as it would offer only the same security as the block cipher itself.

New Finite-Time and Fast Converging Observers With a Single Delay

Abstract: We provide new reduced order observer designs for a key class of nonlinear dynamics. When continuous output measurements are available, we prove that our observers converge in a fixed finite time in the absence of perturbations, and we prove a robustness result under uncertainties in the output measurements and in the dynamics, which bounds the observation error in terms of bounds on the uncertainties. The observers contain a dynamic extension with only one pointwise delay, and they use the observability Gramian to eliminate an invertibility condition that was present in earlier finite time observer designs. We also provide analogs for cases where the measurements are only available at discrete times, where we prove exponential input-to-state stability. We illustrate the advantages of our new observers using a DC motor dynamics.

BrumiR: A toolkit for <i>de novo</i> discovery of microRNAs from sRNA-seq data

Abstract: Abstract MicroRNAs (miRNAs) are small noncoding RNAs that are key players in the regulation of gene expression. In the past decade, with the increasing accessibility of high-throughput sequencing technologies, different methods have been developed to identify miRNAs, most of which rely on preexisting reference genomes. However, when a reference genome is absent or is not of high quality, such identification becomes more difficult. In this context, we developed BrumiR, an algorithm that is able to discover miRNAs directly and exclusively from small RNA (sRNA) sequencing (sRNA-seq) data. We benchmarked BrumiR with datasets encompassing animal and plant species using real and simulated sRNA-seq experiments. The results demonstrate that BrumiR reaches the highest recall for miRNA discovery, while at the same time being much faster and more efficient than the state-of-the-art tools evaluated. The latter allows BrumiR to analyze a large number of sRNA-seq experiments, from plants or animal species. Moreover, BrumiR detects additional information regarding other expressed sequences (sRNAs, isomiRs, etc.), thus maximizing the biological insight gained from sRNA-seq experiments. Additionally, when a reference genome is available, BrumiR provides a new mapping tool (BrumiR2reference) that performs an a posteriori exhaustive search to identify the precursor sequences. Finally, we also provide a machine learning classifier based on a random forest model that evaluates the sequence-derived features to further refine the prediction obtained from the BrumiR-core. The code of BrumiR and all the algorithms that compose the BrumiR toolkit are freely available at https://github.com/camoragaq/BrumiR.

Knowledge extraction from the learning of sequences in a long short term memory (LSTM) architecture

Abstract: Transparency and trust in machine learning algorithms have been deemed to be fundamental and yet, from a practical point of view, they remain difficult to implement. Particularly, explainability and interpretability are certainly among the most difficult capabilities to be addressed and imply to be able to understand a decision in terms of simple cues and rules. In this article, we address this specific problem in the context of sequence learning by recurrent neuronal models (and more specifically Long Short Term Memory model). We introduce a general method to extract knowledge from the latent space based on the clustering of the internal states. From these hidden states, we explain how to build and validate an automaton that corresponds to the underlying (unknown) grammar, and allows to predict if a given sequence is valid or not. Finally, we show that it is possible for such complex recurrent model, to extract the knowledge that is implicitly encoded in the sequences and we report a high rate of recognition of the sequences extracted from the original grammar. This method is illustrated on artificial grammars (Reber grammar variants) as well as on a real use-case in the electrical domain, whose underlying grammar is unknown.

Statistical Decoding 2.0: Reducing Decoding to LPN

Abstract: The security of code-based cryptography relies primarily on the hardness of generic decoding with linear codes. The best generic decoding algorithms are all improvements of an old algorithm due to Prange: they are known under the name of information set decoders (ISD). A while ago, a generic decoding algorithm which does not belong to this family was proposed: statistical decoding. It is a randomized algorithm that requires the computation of a large set of parity-checks of moderate weight, and uses some kind of majority voting on these equations to recover the error. This algorithm was long forgotten because even the best variants of it performed poorly when compared to the simplest ISD algorithm. We revisit this old algorithm by using parity-check equations in a more general way. Here the parity-checks are used to get LPN samples with a secret which is part of the error and the LPN noise is related to the weight of the parity-checks we produce. The corresponding LPN problem is then solved by standard Fourier techniques. By properly choosing the method of producing these low weight equations and the size of the LPN problem, we are able to outperform in this way significantly information set decoders at code rates smaller than 0.3. It gives for the first time after 60 years, a better decoding algorithm for a significant range which does not belong to the ISD family.

The Security Protocol Verifier ProVerif and its Horn Clause Resolution Algorithm

Abstract: ProVerif is a widely used security protocol verifier. Internally, ProVerif uses an abstract representation of the protocol by Horn clauses and a resolution algorithm on these clauses, in order to prove security properties of the protocol or to find attacks. In this paper, we present an overview of ProVerif and discuss some specificities of its resolution algorithm, related to the particular application domain and the particular clauses that ProVerif generates. This paper is a short summary that gives pointers to publications on ProVerif in which the reader will find more details.