Biology subjects

Gaudreault, N.

Publications and source records attributed to Gaudreault, N..

15 recordsLinked to original sources

bioRxiv2026

Trustworthy in silico labeling via semantic visual interpretability of image-to-image translation

Cross-modality image translation promises to provide multiple layers of biological information from a single image input, yet its practical application is stalled by a lack of interpretability and the inability to account for model imperfections. In silico labeling, the inference of organelle localization from label-free images, is a primary example where this black-box nature limits adoption. We present Mask Interpreter, a generalized method for semantic visual interpretability of image-to-image translation models. By uncovering organelle-specific "explanation signatures", Mask Interpreter validates that models rely on authentic biological structures rather than spurious artifacts. Beyond biological validation, it outperforms traditional explainable AI (xAI) approaches, identifies batch effects and localizes prediction errors when ground-truth fluorescence is unavailable. Semantic confidence modeling further provides fine-grained reliability assessment at single-cell resolution, enabling the automated exclusion of artifacts from downstream analyses. By bridging the gap between computational inference and meaningful biological features, Mask Interpreter transforms in silico labeling into a reliable tool for scientific discovery across diverse biomedical imaging modalities.

cell biology↗

bioRxiv2026

Decoding causal genes and programs from regulatory variants in aortic valve disease

Aortic valve disease is common, yet its regulatory mechanisms remain poorly understood. We performed multi-omic profiling of human aortic valve interstitial cells (HAVICs), identifying 11,891 allele-specific chromatin accessibility QTLs (as-caQTLs), 48% novel to this cell type. These variants were enriched in active enhancers, disrupted transcription factor (TF) motifs, particularly AP-1, TEAD and GATA families, and were validated by allele-specific TF binding assays. A fine-tuned deep DNA sequence model prioritized common and rare variants at risk loci predicted to impact chromatin accessibility. Single-cell CRISPRi perturbation of 247 variants identified cis-target genes at 55 as-caQTL elements, including loci without eQTLs. We demonstrate that common regulatory variants controlling elastin and fibrillin impact the development of the aortic valve apparatus. We provide genetic evidence and a mechanistic framework for the contribution of a reduced aortic root size to CAVD risk. Perturbations identified core cell programs led by upstream regulators AHNAK, PDIA6, and RNFT1 converging on extracellular matrix production and iron transport.

genetics↗

bioRxiv2024

Interpretable representation learning for 3D multi-piece intracellular structures using point clouds

A key challenge in understanding subcellular organization is quantifying interpretable measurements of intracellular structures with complex multi-piece morphologies in an objective, robust and generalizable manner. Here we introduce a morphology-appropriate representation learning framework that uses 3D rotation invariant autoencoders and point clouds. This framework is used to learn representations of complex multi-piece morphologies that are independent of orientation, compact, and easy to interpret. We apply our framework to intracellular structures with punctate morphologies (e.g. DNA replication foci) and polymorphic morphologies (e.g. nucleoli). We systematically compare our framework to image-based autoencoders across several intracellular structure datasets, including a synthetic dataset with pre-defined rules of organization. We explore the trade-offs in the performance of different models by performing multi-metric benchmarking across efficiency, generative capability, and representation expressivity metrics. We find that our framework, which embraces the underlying morphology of multi-piece structures, facilitates the unsupervised discovery of sub-clusters for each structure. We show how our approach can also be applied to phenotypic profiling using a dataset of nucleolar images following drug perturbations. We implement and provide all representation learning models using CytoDL, a python package for flexible and configurable deep learning experiments.

cell biology↗

bioRxiv2024

Single-cell and single-nucleus RNA-sequencing from paired normal-adenocarcinoma lung samples provides both common and discordant biological insights.

Whether single-cell RNA-sequencing (scRNA-seq) captures the same biological information as single-nucleus RNA-sequencing (snRNA-seq) remains uncertain and likely to be context-dependent. Herein, a head-to-head comparison was performed in matched normal-adenocarcinoma human lung samples to assess biological insights derived from scRNA-seq versus snRNA-seq and better understand the cellular transition that occurs from normal to tumoral tissue. Here, the transcriptome of 160,621 cells/nuclei was obtained. In non-tumor lung, cell type proportions varied widely between scRNA-seq and snRNA-seq with a predominance of immune cells in the former (81.5%) and epithelial cells (69.9%) in the later. Similar results were observed in adenocarcinomas, in addition to an overall increase in cell type heterogeneity and a greater prevalence of copy number variants in cells of epithelial origin, which suggests malignant assignment. The cell type transition that occurs from normal lung tissue to adenocarcinoma was not always concordant whether cells or nuclei were examined. As expected, large differential expression of the whole-cell and nuclear transcriptome was observed, but cell-type specific changes of paired normal and tumor lung samples revealed a set of common genes in the cells and nuclei involved in cancer-related pathways. In addition, we showed that the ligand-receptor interactome landscape of lung adenocarcinoma was largely different whether cells or nuclei were evaluated. Immune cell depletion in fresh specimens partly mitigated the difference in cell type composition observed between cells and nuclei. However, the extra manipulations affected cell viability and amplified the transcriptional signatures associated with stress responses. In conclusion, research applications focussing on mapping the immune landscape of lung adenocarcinoma benefit from scRNA-seq in fresh samples, whereas snRNA-seq of frozen samples provide a low-cost alternative to profile more epithelial and cancer cells, and yield cell type proportions that more closely match tissue content. Author SummarySingle-cell transcriptomic datasets provide unprecedented opportunities to disentangle the complex tissue microenvironment and cellular origin of cancer. Data are scarce regarding the pros and cons of single-cell RNA sequencing (scRNA-seq) of freshly explanted human tissues over single-nuclei sequencing (snRNA-seq) from the same archived frozen tissues. Lung adenocarcinoma represents a medically valuable case study to compare the biological signal recovered through cells and nuclei sequencing. Here, we sequenced the transcriptome of 160,621 cells/nuclei in paired normal-adenocarcinoma lung samples. Cell type proportions varied widely between scRNA-seq and snRNA-seq with a predominance of immune cells in the former and epithelial cells in the later. Adenocarcinomas were characterized by an increase in cell type heterogeneity and a greater prevalence of malignant epithelial cells in both scRNA-seq and snRNA-seq. The cellular and gene expression transition that occur from normal lung to adenocarcinoma showed common and discordant biological insights whether cells or nuclei were examined. Research applications focussing on mapping the immune landscape of lung cancer benefit from scRNA-seq in fresh samples, whereas snRNA-seq of the same frozen samples provide a low-cost and more flexible alternative to profile more epithelial and cancer cells, and yield cell type proportions that more closely match tissue content.

cancer biology↗

bioRxiv2021

A deep generative model of 3D single-cell organization

1 AbstractWe introduce a framework for end-to-end integrative modeling of 3D single-cell multi-channel fluorescent image data of diverse subcellular structures. We employ stacked conditional {beta}-variational autoencoders to first learn a latent representation of cell morphology, and then learn a latent representation of subcellular structure localization which is conditioned on the learned cell morphology. Our model is flexible and can be trained on images of arbitrary subcellular structures and at varying degrees of sparsity and reconstruction fidelity. We train our full model on 3D cell image data and explore design trade-offs in the 2D setting. Once trained, our model can be used to impute structures in cells where they were not imaged and to quantify the variation in the location of all subcellular structures by generating plausible instantiations of each structure in arbitrary cell geometries. We apply our trained model to a small drug perturbation screen to demonstrate its applicability to new data. We show how the latent representations of drugged cells differ from unperturbed cells as expected by on-target effects of the drugs. 2 Author summaryIts impossible to acquire all the information we want about every cell were interested in in a single experiment. Even just limiting ourselves to imaging, we can only image a small set of subcellular structures in each cell. If we are interested in integrating those images into a holistic picture of cellular organization directly from data, there are a number of approaches one might take. Here, we leverage the fact that of the three channels we image in each cell, two stay the same across the data set; these two channels assess the cells shape and nuclear morphology. Given these two reference channels, we learn a model of cell and nuclear morphology, and then use this as a reference frame in which to learn a representation of the localization of each subcellular structure as measured by the third channel. We use {beta}-variational autoencoders to learn representations of both the reference channels and representations of each subcellular structure (conditioned on the reference channels of the cell in which it was imaged). Since these models are both probabilistic and generative, we can use them to understand the variation in the data from which they were trained, to generate instantiations of new cell morphologies, and to generate imputations of structures in real cell images to create an integrated model of subcellular organization.

cell biology↗

bioRxiv2021

Towards community-driven metadata standards for light microscopy: tiered specifications extending the OME model

1 -Digital light microscopy provides powerful tools for quantitatively probing the real-time dynamics of subcellular structures. While the power of modern microscopy techniques is undeniable, rigorous record-keeping and quality control are required to ensure that imaging data may be properly interpreted (quality), reproduced (reproducibility), and used to extract reliable information and scientific knowledge which can be shared for further analysis (value). Keeping notes on microscopy experiments and quality control procedures ought to be straightforward, as the microscope is a machine whose components are defined and the performance measurable. Nevertheless, to this date, no universally adopted community-driven specifications exist that delineate the required information about the microscope hardware and acquisition settings (i.e., microscopy "data provenance" metadata) and the minimally accepted calibration metrics (i.e., microscopy quality control metadata) that should be automatically recorded by both commercial microscope manufacturers and customized microscope developers. In the absence of agreed guidelines, it is inherently difficult for scientists to create comprehensive records of imaging experiments and ensure the quality of resulting image data or for manufacturers to incorporate standardized reporting and performance metrics. To add to the confusion, microscopy experiments vary greatly in aim and complexity, ranging from purely descriptive work to complex, quantitative and even sub-resolution studies that require more detailed reporting and quality control measures. To solve this problem, the 4D Nucleome Initiative (4DN) (1, 2) Imaging Standards Working Group (IWG), working in conjunction with the BioImaging North America (BINA) Quality Control and Data Management Working Group (QC-DM-WG) (3), here propose light Microscopy Metadata specifications that scale with experimental intent and with the complexity of the instrumentation and analytical requirements. They consist of a revision of the Core of the Open Microscopy Environment (OME) Data Model, which forms the basis for the widely adopted Bio-Formats library (4-6), accompanied by a suite of three extensions, each with three tiers, allowing the classification of imaging experiments into levels of increasing imaging and analytical complexity (7, 8). Hence these specifications not only provide an OME-based comprehensive set of metadata elements that should be recorded, but they also specify which subset of the full list should be recorded for a given experimental tier. In order to evaluate the extent of community interest, an extensive outreach effort was conducted to present the proposed metadata specifications to members of several core-facilities and international bioimaging initiatives including the European Light Microscopy Initiative (ELMI), Global BioImaging (GBI), and European Molecular Biology Laboratory (EMBL) - European Bioinformatics Institute (EBI). Consequently, close ties were established between our endeavour and the undertakings of the recently established QUAlity Assessment and REProducibility for Instruments and Images in Light Microscopy global community initiative (9). As a result this flexible 4DN-BINA-OME (NBO namespace) framework (7, 8) represents a turning point towards achieving community-driven Microscopy Metadata standards that will increase data fidelity, improve repeatability and reproducibility, ease future analysis and facilitate the verifiable comparison of different datasets, experimental setups, and assays, and it demonstrates the method for future extensions. Such universally accepted microscopy standards would serve a similar purpose as the Encode guidelines successfully adopted by the genomic community (10, 11). The intention of this proposal is therefore to encourage participation, critiques and contributions from the entire imaging community and all stakeholders, including research and imaging scientists, facility personnel, instrument manufacturers, software developers, standards organizations, scientific publishers, and funders.

bioinformatics↗

bioRxiv2020

Automated hiPSC culture and sample preparation for 3D live cell microscopy

Our goal is to identify and understand cellular behaviors using 3D live imaging of cell organization. To do this, we image human inducible pluripotent stem cell (hiPSC) lines expressing fluorescently tagged protein representing specific cellular organelles and structures. To produce large numbers of standardized cell images, we developed an automated hiPSC culture procedure, to maintain, passage and Matrigel coat 6-well plastic plates and 96-well glass plates compatible with high-resolution 3D microscopy. Here we describe this system including optimization procedures and specific values for plate movement, angle of tips, speed of aspiration and dispense, seeding strategies and timing of every step. We validated this approach through a side-by-side comparison of quality control results obtained from manual and automated methods. Additionally, we developed an automated image-based colony segmentation and feature extraction pipeline to predict cell count and select wells with consistent morphology for high resolution 3D microscopy.

cell biology↗

bioRxiv2020

Cutaneous inputs from perineal region facilitates and modulates spinal locomotor activity and reduces cutaneous reflexes from the foot in spinal cats

It is well known that mechanically stimulating the perineal region potently facilitates hindlimb locomotion and weight support in mammals with a spinal transection (spinal mammals). However, how perineal stimulation mediates this excitatory effect is poorly understood. We evaluated the effect of mechanically stimulating (vibration or pinch) the perineal region on ipsilateral (9-14 ms onset) and contralateral (14-18 ms onset) short-latency cutaneous reflex responses evoked by electrically stimulating the superficial peroneal or distal tibial nerve in seven adult spinal cats where hindlimb movement was restrained. Cutaneous reflexes were evoked before, during, and after mechanical stimulation of the perineal region. We found that vibration or pinch of the perineal region effectively triggered rhythmic activity, unilateral and bilateral to nerve stimulation. When electrically stimulating nerves, adding perineal stimulation modulated rhythmic activity by decreasing cycle and burst durations and by increasing the amplitude of flexors and extensors. Perineal stimulation also disrupted the timing of the ipsilateral rhythm, which had been entrained by nerve stimulation. Mechanically stimulating the perineal region decreased ipsilateral and contralateral short-latency reflex responses evoked by cutaneous inputs, a phenomenon we observed in muscles crossing different joints and located in different limbs. The results suggest that the excitatory effect of perineal stimulation on locomotion and weight support is not mediated by increasing cutaneous reflex gain and instead points to an excitation of central pattern-generating circuitry. Our results are consistent with a state-dependent modulation of reflexes by spinal interneuronal circuits. Significance StatementMechanically stimulating the skin of the perineal region strongly facilitates hindlimb locomotion in mammals following a complete spinal cord injury (SCI). Despite its remarkable effectiveness in promoting hindlimb locomotion in spinal cord-injured mammals, we do not know how this is mediated. The present study provides data on how inputs from the perineal region interact with neuronal circuits that generate locomotor-like activity and reflexes from the foot. A better understanding of how inputs from the perineal region interact with neuronal circuits of the spinal cord could lead to non-invasive approaches to restore walking in people with SCI.

neuroscience↗

bioRxiv2020

Cell states beyond transcriptomics: integrating structural organization and gene expression in hiPSC-derived cardiomyocytes

We present a quantitative co-analysis of RNA abundance and sarcomere organization in single cells and an integrated framework to predict subcellular organization states from gene expression. We used human induced pluripotent stem cell (hiPSC)-derived cardiomyocytes expressing mEGFP-tagged alpha-actinin-2 to develop quantitative image analysis tools for systematic and automated classification of subcellular organization. This captured a wide range of sarcomeric organization states within cell populations that were previously difficult to quantify. We performed RNA FISH targeting genes identified by single cell RNA sequencing to simultaneously assess the relationship between transcript abundance and structural states in single cells. Co-analysis of gene expression and sarcomeric patterns in the same cells revealed biologically meaningful correlations that could be used to predict organizational states. This study establishes a framework for multi-dimensional analysis of single cells to study the relationships between gene expression and subcellular organization and to develop a more nuanced description of cell states. Graphical AbstractTranscriptional profiling and structural classification was performed on human induced pluripotent stem cell-derived cardiomyocytes to characterize the relationship between transcript abundance and subcellular organization. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=199 SRC="FIGDIR/small/081083v1_ufig1.gif" ALT="Figure 1"> View larger version (60K): org.highwire.dtl.DTLVardef@8f89c4org.highwire.dtl.DTLVardef@19dc813org.highwire.dtl.DTLVardef@1ba7c20org.highwire.dtl.DTLVardef@2b3daa_HPS_FORMAT_FIGEXP M_FIG C_FIG

cell biology↗

bioRxiv2019

Genetic association analyses highlight IL6, ALPL, and NAV1 as three new susceptibility genes underlying calcific aortic valve stenosis

To date, only two replicated loci, LPA and PALMD, have been identified as causal genes for calcific aortic valve stenosis (CAVS) using genome-wide and transcriptome-wide association study (TWAS). To identify additional susceptibility genes for CAVS, we performed a GWAS meta-analysis totaling 5,115 cases and 354,072 controls of European descent. Four loci achieved genome-wide significance, including two new loci: IL6 (interleukin 6) on 7p15.3 and ALPL (alkaline phosphatase) on 1p36.12. A TWAS integrating an eQTL study of 233 human aortic valves identified NAV1 (neuron navigator 1) on 1q32.1 as a new candidate causal gene. The CAVS risk alleles were associated with higher mRNA expression of NAV1 in valve tissues. Association results at the genome-wide scale showed genetic correlation with coronary artery disease and cardiovascular risk factors. Our study highlights three new loci implicating inflammation, mineralization and blood vessel integrity in CAVS pathogenesis and supports shared genetic etiology with cardiovascular traits.

genomics↗

bioRxiv2018

Biological variation in the sizes, shapes and locations of visual cortical areas in the mouse

Visual cortex is organized into discrete sub-regions or areas that are arranged into a hierarchy and serve different functions in the processing of visual information. In our previous work, we noted that retinotopic maps of cortical visual areas differed between mice, but did not quantify these differences or determine the relative contributions of biological variation and measurement noise. Here we quantify the biological variation in the size, shape and locations of 11 visual areas in the mouse. We find that there is substantial biological variation in the sizes of visual areas, with some visual areas varying in size by two-fold across the population of mice.

neuroscience↗

bioRxiv2018

Classification of electrophysiological and morphological types in mouse visual cortex

Understanding the diversity of cell types in the brain has been an enduring challenge and requires detailed characterization of individual neurons in multiple dimensions. To profile morpho-electric properties of mammalian neurons systematically, we established a single cell characterization pipeline using standardized patch clamp recordings in brain slices and biocytin-based neuronal reconstructions. We built a publicly-accessible online database, the Allen Cell Types Database, to display these data sets. Intrinsic physiological and morphological properties were measured from over 1,800 neurons from the adult laboratory mouse visual cortex. Quantitative features were used to classify neurons into distinct types using unsupervised methods. We establish a taxonomy of morphologically- and electrophysiologically-defined cell types for this region of cortex with 17 e-types and 35 m-types, as well as an initial correspondence with previously-defined transcriptomic cell types using the same transgenic mouse lines.

neuroscience↗

bioRxiv2018

A large-scale, standardized physiological survey reveals higher order coding throughout the mouse visual cortex

To understand how the brain processes sensory information to guide behavior, we must know how stimulus representations are transformed throughout the visual cortex. Here we report an open, large-scale physiological survey of neural activity in the awake mouse visual cortex: the Allen Brain Observatory Visual Coding dataset. This publicly available dataset includes cortical activity from nearly 60,000 neurons collected from 6 visual areas, 4 layers, and 12 transgenic mouse lines from 221 adult mice, in response to a systematic set of visual stimuli. Using this dataset, we reveal functional differences across these dimensions and show that visual cortical responses are sparse but correlated. Surprisingly, responses to different stimuli are largely independent, e.g. whether a neuron responds to natural scenes provides no information about whether it responds to natural movies or to gratings. We show that these phenomena cannot be explained by standard local filter-based models, but are consistent with multi-layer hierarchical computation, as found in deeper layers of standard convolutional neural networks.

neuroscience↗

bioRxiv2018

The organization of intracortical connections by layer and cell class in the mouse brain

The mammalian cortex is a laminar structure composed of many cell types densely interconnected in complex ways. Recent systematic efforts to map the mouse mesoscale connectome provide comprehensive projection data on interareal connections, but not at the level of specific cell classes or layers within cortical areas. We present here a significant expansion of the Allen Mouse Brain Connectivity Atlas, with [~]1,000 new axonal projection mapping experiments across nearly all isocortical areas in 49 Cre driver lines. Using 13 lines selective for cortical layer-specific projection neuron classes, we identify the differential contribution of each layer/class to the overall intracortical connectivity patterns. We find layer 5 (L5) projection neurons account for essentially all intracortical outputs. L2/3, L4, and L6 neurons contact a subset of the L5 cortical targets. We also describe the most common axon lamination patterns in cortical targets. Most patterns are consistent with previous anatomical rules used to determine hierarchical position between cortical areas (feedforward, feedback), with notable exceptions. While diverse target lamination patterns arise from every source layer/class, L2/3 and L4 neurons are primarily associated with feedforward type projection patterns and L6 with feedback. L5 has both feedforward and feedback projection patterns. Finally, network analyses revealed a modular organization of the intracortical connectome. By labeling interareal and intermodule connections as feedforward or feedback, we present an integrated view of the intracortical connectome as a hierarchical network.

neuroscience↗

bioRxiv2017

A transcriptome-wide association study identifies PALMD as a susceptibility gene for calcific aortic valve stenosis

Calcific aortic valve stenosis (CAVS) is a common and life-threatening heart disease with no drug that can stop or delay its progression. Elucidating the genetic factors underpinning CAVS is an urgent priority to find new therapeutic targets1. Major landmarks in genetics of CAVS include the discoveries of NOTCH12 and LPA3. However, genetic variants in these genes accounted for a small number of cases and low population-attributable risk. Here we mapped a new susceptibility locus for CAVS on chromosome 1p21.2 and identified PALMD (palmdelphin) as the causal gene. PALMD was revealed using a transcriptome-wide association study (TWAS)4, which combines a genome-wide association study (GWAS) of 1,009 cases and 1,017 ethnically-matched controls with the first large-scale expression quantitative trait loci (eQTL) mapping study on human aortic valve tissues (n=233). The CAVS risk alleles and increasing disease severity were both associated with lowered mRNA expression levels of PALMD in valve tissues. The top variant explained up to 12.5% of the population-attributable risk and showed similar effect and strong association with CAVS (P=1.53 x 10-10) in UK Biobank comparing 1,391 cases and 352,195 controls. The identification of PALMD as a susceptibility gene for CAVS provides new insights about the genetic nature of this disease and opens new avenues to investigate its etiology and develop much-needed therapeutic options.

genomics↗