Biology subjects

Pelletier, D.

Publications and source records attributed to Pelletier, D..

5 recordsLinked to original sources

bioRxiv2022

Probabilistic Brain Extraction in MR Images via Conditional Generative Adversarial Networks

Brain extraction, which refers to the task of segmenting brain tissue in an MR image of a subject, forms an essential first step for many quantitative neuroimaging applications. These include quantifying grey and white matter volumes, monitoring neurological diseases like multiple sclerosis (MS) and Alzheimers disease, and estimating brain atrophy. Over the years several algorithms that automate brain extraction have been proposed. More recently, novel image-to-image deep learning methods have been implemented for this task, and have demonstrated significant gains in accuracy and robustness. However, to our knowledge, none of these algorithms account for the uncertainty that is inherent in brain extraction. Motivated by this, we propose a novel, probabilistic deep learning algorithm for brain extraction that recasts this task as a Bayesian inference problem, and then utilizes a conditional generative adversarial network (cGAN) to solve it. The input to the generator network is an MR image of the head and the output is a collection of images of the brain that are drawn from a probability density conditioned on the input image. These images are used to generate a pixel-wise mean image, which serves as the best guess for an image of the brain, and a pixel-wise standard deviation image, which quantifies the uncertainty in the prediction. We test this algorithm on head MR images of fifty subjects and demonstrate that it is more accurate than a commonly used brain extraction tool, and that its performance compares well with the current state of the art in deep learning algorithms. We also demonstrate the utility of the estimates of uncertainty generated by the algorithm.

bioengineering↗

bioRxiv2020

Host auxin machinery is important in mediating root architecture changes induced by Pseudomonas fluorescens GM30 and Serendipita indica

Auxin is a key phytohormone that is integral to plant developmental processes including those underlying root initiation, elongation, and branching. Beneficial microbes have been shown to have an impact on root development, potentially mediated through auxin. In this study, we explore the role of host auxin signaling and transport components in mediating the root growth promoting effects of beneficial microbes. Towards this end, we undertook co-culture studies of Arabidopsis thaliana plants with microbes previously reported to promote lateral root proliferation and produce auxin. Two types of beneficial microbes were included in the present study; a plant growth promoting bacterial species of interest, Pseudomonas fluorescens GM30, and a well-studied plant growth promoting fungal species, Serendipita indica (Piriformospora indica). Following co-culture, lateral root production was found restored in auxin transport inhibitor-treated plants, suggesting involvement of microbe and/or microbially-produced auxin in altering plant auxin levels. In order to clarify the role of host auxin signaling and transport pathways in mediating interactions with bacterial and fungal species, we employed a suite of auxin genetic mutants as hosts in co-culture screens. Our result show that the transport proteins PIN2, PIN3, and PIN7 and the signaling protein ARF19, are required for mediating root architecture effects by the bacterial and/or fungal species. Mutants corresponding to these proteins did not significantly respond to co-culture treatment and did not show increases in lateral root production and lateral root density. These results implicate the importance of host auxin signaling in both bacterial and fungal induced changes in root architecture and serve as a driver for future research on understanding the role of auxin-dependent and auxin-independent pathways in mediating plant-microbe interactions in economically important crop species.

plant biology↗

bioRxiv2020

Marine reserve benefits and recreational fishing yields: the winners and the losers

Marine reserves constitute effective tools for preserving fish stocks and associated human benefits. However, not all reserves perform equally, and predicting the response of marine communities to management actions in the long run is challenging. Our decadal-scale survey of recreational fishing yields at Frances 45-year old Cerbere-Banyuls marine reserve indicated significant protection benefits, with 40-50% higher fishing yields per unit effort in the partial-protection zone of the reserve (where fishing is permitted but at a lower level) than in surrounding non-reserve areas. Over the period 2005-2014, catch per unit effort (CPUE) showed a similar pattern of decline inside and outside the reserve while weight per unit effort (WPUE) increased by 131% inside and decreased by 60% outside. Different CPUE and WPUE trajectories among fish families indicated changing catch assemblages, with yields increasing for the family most valued by fisheries, Sparidae (the ecological winners). However, reserve benefits were restricted to off-shore fishermen (the social winners), as on-shore yields were [~]4 times lower and declining, even inside the reserve. Our study illustrates how surveys of recreational fishing yields can help evaluate the effectiveness of marine protected areas for key social and ecological protagonists. We show that, more than four decades after its establishment, fishing efficiencies at the historical Cerbere-Banyuls marine reserve are still changing, but benefits in terms of catch abundance, weight, and composition remain predominantly restricted to off-shore fishermen. Further regulations appear necessary to guarantee that conservation strategies equitably benefit societal groups.

ecology↗

bioRxiv2019

Crystal structure of human PACRG in complex with MEIG1

In human, the Parkin Co-Regulated Gene (PACRG) shares a bidirectional promoter with Parkin, a gene involved in Parkinsons disease, mitochondrial quality control and inflammation. The PACRG protein is essential to the formation of the inner junction between doublet microtubules of the axoneme, a structure found in flagella and cilia. PACRG interacts with tubulin as well as the meiosis expressed gene 1 (MEIG1) protein, which is essential for spermiogenesis in mice. However, the 3D structure of human PACRG is unknown. Here, we report the crystal structure of the C-terminal domain of human PACRG in complex with MEIG1 at 2.1 [A] resolution. PACRG adopts an -helical structure with a loop insertion that mediates a conserved network of interactions with MEIG1. Using the cryo-electron tomography structure of the axonemal doublet microtubule from the flagellated protozoan Chlamydomonas reinhardtii, we generate a model of a mammalian microtubule doublet inner junction, which reveals how PACRG interacts with tubulin subunits in both the A- and B-tubules. Furthermore, the model shows that MEIG1 interacts with {beta}-tubulin on the outer surface of the B-tubule, facing towards the central pair of the axoneme. We also model the PACRG-like protein (PACRGL), a homolog of PACRG with potential roles in microtubule remodelling and axonemal inner junction formation. Finally, we explore the evolution of the PACRG and Parkin head-to-head gene structure and analyze the tissue distribution of their transcripts. Our work establishes a framework to assess the function of the PACRG family of proteins and its adaptor proteins in the function of motile and non-motile cilia.

cell biology↗

bioRxiv2017

An automated statistical technique for counting distinct multiple sclerosis lesions can recover aspects of lesion history and provide relevant disease information

BackgroundLesion load is a common biomarker in multiple sclerosis, yet it has historically shown modest associations with clinical outcomes. Lesion count, which encapsulates the natural history of lesion formation and is thought to provide complementary information, is difficult to assess in patients with confluent (i.e. spatially overlapping) lesions. We introduce a statistical technique for cross-sectionally counting pathologically distinct lesions.\n\nMethodsMRI is used to assess the probability of lesion at each location. The texture of this map is quantified using a novel technique, and clusters resembling the center of a lesion are counted.\n\nResultsValidity was demonstrated by comparing the proposed count to a gold-standard count in 60 subjects observed longitudinally. The counts were highly correlated (r = .97, p < .001) and not significantly different (t59 = -0.83, p > .40). Reliability was determined using 14 scans of a clinically stable subject acquired at 7 sites, and variability of lesion count was equivalent to that of lesion load. Accounting for lesion load and age, lesion count was negatively associated (t58 = -2.73, p < .01) with the Expanded Disability Status Scale (EDSS). Average lesion size had a higher association with EDSS (r =.35, p < .01) than lesion load (r = .10, p > .40) or lesion count (r = -.12, p > .30) alone.\n\nConclusionThese findings demonstrate that it is possible to recover important aspects of the natural history of lesion formation without longitudinal data, and suggest that lesion size provides complementary information about disease.\n\nGrant SupportThe project described was supported in part by the NIH grants R01 NS085211, R21 NS093349, and R01 NS094456 from the National Institute of Neurological Disorders and Stroke (NINDS). The study was also supported by the Intramural Research Program of NINDS and the Race to Erase MS Foundation. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding agencies.

neuroscience↗