Biology subjects

Drouin, A.

Publications and source records attributed to Drouin, A..

5 recordsLinked to original sources

bioRxiv2026

Status of Round Goby Invasion Fronts in New York and Quebec: Implications for Lake Champlain

Invasive round goby Neogobius melanostomus have advanced eastward through the state of New York and provinces of Ontario and Quebec over the past two decades and are approaching Lake Champlain, one of the largest lakes in North America. This manuscript describes international efforts to monitor round goby populations during 2021-2025 on (a) the southern approach to Lake Champlain via the Hudson River and Champlain Canal, and (b) the northern approach to Lake Champlain via the Saint Lawrence River and Richelieu River. Monitoring utilized environmental DNA (eDNA), backpack electrofishing, beach seining, benthic trawling, and viral hemorrhagic septicemia virus (VHSV) testing. In the Champlain Canal, round goby were captured as far north as the downstream side of the C1 dam (97 kilometers [km] from Lake Champlain) while eDNA detections occurred as far north as the upstream side of the C2 dam (90 km from Lake Champlain). In the Richelieu River, round goby were captured as far south as Saint-Marc-sur-Richelieu (82 km from Lake Champlain) while the southern-most eDNA detections occurred near the Canadian side of the international border (4 km from Lake Champlain). Water temperature influenced habitat usage of round goby in the Champlain Canal, with catch rates in near-shore areas declining at < 10 {degrees}C. All VHSV test results were non-detections at the mouth of the Richelieu River, while one positive and two inconclusive results occurred along the Champlain Canal. Together, these data have informed multiple mitigation measures and have implications for management of aquatic invasive species across North America.

ecology↗

bioRxiv2024

NeuroTorch: A Python library for neuroscience-oriented machine learning

Machine learning (ML) has become a powerful tool for data analysis, leading to significant advances in neuroscience research. While ML algorithms are proficient in general-purpose tasks, their highly technical nature often hinders their compatibility with the observed biological principles and constraints in the brain, thereby limiting their suitability for neuroscience applications. In this work, we introduce NeuroTorch, a comprehensive ML pipeline specifically designed to assist neuroscientists in leveraging ML techniques using biologically inspired neural network models. NeuroTorch enables the training of recurrent neural networks equipped with either spiking or firing-rate dynamics, incorporating additional biological constraints such as Dales law and synaptic excitatory-inhibitory balance. The pipeline offers various learning methods, including backpropagation through time and eligibility trace forward propagation, aiming to allow neuroscientists to effectively employ ML approaches. To evaluate the performance of NeuroTorch, we conducted experiments on well-established public datasets for classification tasks, namely MNIST, Fashion-MNIST, and Heidelberg. Notably, NeuroTorch achieved accuracies that replicated the results obtained using the Norse and SpyTorch packages. Additionally, we tested NeuroTorch on real neuronal activity data obtained through volumetric calcium imaging in larval zebrafish. On training sets representing 9.3 minutes of activity under darkflash stimuli from 522 neurons, the mean proportion of variance explained for the spiking and firing-rate neural network models, subject to Dales law, exceeded 0.97 and 0.96, respectively. Our analysis of networks trained on these datasets indicates that both Dales law and spiking dynamics have a beneficial impact on the resilience of network models when subjected to connection ablations. NeuroTorch provides an accessible and well-performing tool for neuroscientists, granting them access to state-of-the-art ML models used in the field without requiring in-depth expertise in computer science.

neuroscience↗

bioRxiv2022

Environmental determinants of invasion hotspots and refugia at a river scale: implications for conservation of native biodiversity

Introductions of exotic invasive species are a global disturbance for natural habitats. The severity of invasions can greatly vary from local to global scales, as observed in invasion refuges, exhibiting lower-than-expected invasion intensity. In this study, we analyzed the effects of water conductivity and wetland presence on the density of the round goby (Neogobius melanostomus) in a large-scale study (> 1300 sites), spanning a 400 km stretch of the St. Lawrence River (Canada). Our results showed that round goby density was null in sites with water conductivity under 100 {micro}S/cm and increased toward a probable biological optimum at 300 {micro}S/cm. The presence of wetlands appeared to also decrease round goby density along the conductivity continuum. Similarly, fish community diversity was maximal outside of the round goby water conductivity optimum. Hence, low water conductivity (<100 {micro}S/cm), in interaction with the presence of wetlands, can provide a refuge for native aquatic species, establishing a simple risk assessment tool for managers. Our results also highlighted the high value of wetland conservation for conservation of native species biodiversity.

ecology↗

bioRxiv2018

Interpretable genotype-to-phenotype classifiers with performance guarantees

Understanding the relationship between the genome of a cell and its phenotype is a central problem in precision medicine. Nonetheless, genotype-to-phenotype prediction comes with great challenges for machine learning algorithms that limit their use in this setting. The high dimensionality of the data tends to hinder generalization and challenges the scalability of most learning algorithms. Additionally, most algorithms produce models that are complex and difficult to interpret. We alleviate these limitations by proposing strong performance guarantees, based on sample compression theory, for rule-based learning algorithms that produce highly interpretable models. We show that these guarantees can be leveraged to accelerate learning and improve model interpretability. Our approach is validated through an application to the genomic prediction of antimicrobial resistance, an important public health concern. Highly accurate models were obtained for 12 species and 56 antibiotics, and their interpretation revealed known resistance mechanisms, as well as some potentially new ones. An open-source disk-based implementation that is both memory and computationally efficient is provided with this work. The implementation is turnkey, requires no prior knowledge of machine learning, and is complemented by comprehensive tutorials.

genomics↗

bioRxiv2018

Mass Spectra alignment using virtual lock-masses

Mass spectrometry is a valued method to evaluate the metabolomics content of a biological sample. The recent advent of rapid ionization technologies such as Laser Diode Thermal Desorption (LDTD) and Direct Analysis in Real Time (DART) has rendered high-throughput mass spectrometry possible. It is used for large-scale comparative analysis of populations of samples. In practice, many factors resulting from the environment, the protocol, and even the instrument itself, can lead to minor discrepancies between spectra, rendering automated comparative analysis difficult. In this work, a sequence/pipeline of algorithms to correct variations between spectra is proposed. The algorithms correct multiple spectra by identifying peaks that are common to all and, from those, computes a spectrum-specific correction. We show that these algorithms increase comparability within large datasets of spectra, facilitating comparative analysis, such as machine learning.

bioinformatics↗