Biology subjects

McEwen, B.

Publications and source records attributed to McEwen, B..

7 recordsLinked to original sources

bioRxiv2026

BioDCASE: Using data challenges to make community advances in computational bioacoustics

Computational bioacoustics has seen significant advances in recent decades. However, the rate of insights from automated analysis of bioacoustic audio lags behind our rate of collecting the data - due to key capacity constraints in data annotation and bioacoustic algorithm development. Gaps in analysis methodology persist: not because they are intractable, but because of resource limitations in the bioacoustics community. To bridge these gaps, we advocate the open science method of data challenges, structured as public contests. We conducted a bioacoustics data challenge named BioDCASE, within the format of an existing event (DCASE). In this work we report on the procedures needed to select and then conduct useful bioacoustics data challenges. We consider aspects of task design such as dataset curation, annotation, and evaluation metrics. We report the three tasks included in BioDCASE 2025 and the resulting progress made. Based on this we make recommendations for open community initiatives in computational bioacoustics.

animal behavior and cognition↗

bioRxiv2025

Data-driven Sampling Strategies for Fine-Tuning Bird Detection Models

Passive Acoustic Monitoring has emerged as a promising tool for collecting ecological data, particularly in the context of bird population monitoring. Bird species can be automatically identified using pre-trained models, such as BirdNET. The performance of these models can be significantly improved through fine-tuning with annotated samples recorded in the specific acoustic conditions in which the microphones are deployed. However, PAM collects vast amounts of data, and annotating bird vocalizations requires specialized expetise. As a result, only a very small portion of the recordings can be effectively labeled. Selecting the most relevant samples to annotate in order to maximize performance in model fine-tuning remains a significant challenge. First, a regularization technique addresses the challenge of class imbalance during model fine-tuning. Next, a data-driven methodology is developed, introducing the influence score, which quantifies the impact of individual training samples on model performance to inform sampling strategies. A linear model is proposed to estimate the influence score for generalization to unseen data. Finally, several sampling strategies are compared, based on acoustic indices and predictions of the pre-trained model. Together, these contributions enable the identification of efficient annotation strategies to overcome the challenges of limited annotation resources in large-scale passive acoustic monitoring.

bioinformatics↗

bioRxiv2025

Stratified Active Learning for Spatiotemporal Generalisation in Large-Scale Bioacoustic Monitoring

Active learning optimises machine learning model training through the data-efficient selection of informative samples for annotation and training. In the context of biodiversity monitoring using passive acoustic monitoring, active learning offers a promising strategy to reduce the fundamental annotation bottleneck and improve global training efficiency. However, the generalisability of model performance across ecologically relevant strata (e.g. sites, season etc) is often overlooked. As passive acoustic monitoring is extended to larger scales and finer resolutions, inter-strata spatiotemporal variability also increases. We introduce and investigate the concept of stratified active learning to achieve reliable and generalisable model performance across deployment conditions. We compare between implicit cluster-based diversification methods and explicit stratification, demonstrating that cross-strata generalisation is a function of stratum divergence, not sampling balance. Additionally, mutual information as well as exclusion analysis show that spatiotemporal context can explain a substantial proportion of species label variance and inform sampling decisions.

bioinformatics↗

bioRxiv2021

Genomic Modules and Intramodular Network Convergency of Susceptibility and Resilience in Multimodeled Stress in Male Mice

The multifactorial etiology of stress-related disorders is a challenge in developing synchronized medical standards for treatment and diagnosis. It is largely unknown whether there exists molecular convergence in preclinical models of stress generated using disparate construct validity. Using RNA-sequencing (RNA-seq), we investigated the genomic signatures in the ventral hippocampus, which mostly regulates affective behavior, in mouse models that recapitulate the hallmarks of anxiety and depression. Chronic oral corticosterone (CORT), a model that causes a blunted endocrine response to stress, induced anxiety- and depression-like behavior in wildtype mice and mice heterozygous for the gene coding for brain-derived neurotrophic factor (BDNF) Val66Met, a variant associated with genetic susceptibility to stress. In a separate set of mice, chronic social defeat stress led to a susceptible or a resilient population, whose proportion was dependent on housing conditions, standard housing or enriched environment. A rank-rank-hypergeometric (RRHO) analysis of the RNA-seq data set across models demonstrated that in mice treated with CORT and susceptible mice raised in standard housing differentially expressed genes (DEGs) converged toward gene networks involved in similar biological functions. Weighted gene co-expression analysis generated 54 unique modules of interconnected gene hubs, two of which included a combination of all experimental groups and were significantly enriched in DEGs, whose function was consistent with that predicted in the RRHO GO analysis. This multimodel approach showed transcriptional synchrony between models of stress with hormonal, environmental or genetic construct validity shedding light on common genomic drivers that embody the multifaceted nature of stress-related disorders.

neuroscience↗

bioRxiv2019

Metabolic signature in nucleus accumbens for anti-depressant-like effects of acetyl-L-carnitine: An in vivo 1H-magnetic resonance spectroscopy study at 14 T

BackgroundEmerging evidence suggests that hierarchical status may provide vulnerability to develop stress-induced depression. Energy metabolism in the nucleus accumbens (NAc) was recently related to hierarchical status and vulnerability to develop depression-like behavior. Acetyl-L-carnitine (LAC), a mitochondria-boosting supplement, has shown promising antidepressant-like effects opening promising therapeutic strategies for restoring energy balance in depressed patients. Here, we investigated the metabolic impact in the NAc of antidepressant LAC treatment in chronically stressed mice.\n\nMethodMice were characterized for emotional behaviors and social rank. They were then exposed to chronic restraint stress (CRS) for 21 days and subsequently tested in a social behavior (SB) test. A group of mice was also given LAC supplementation during the 7 last CRS days. Mice were then tested in the SB and forced swim tests (FST) and scanned in vivo using 1H-magnetic resonance spectroscopy (1H-MRS) to quantitatively assess the NAc neurochemical profile.\n\nResultsDominant, but not subordinate, mice showed behavioral vulnerability to CRS. In the NAc, dominant mice showed reduced levels of several energy-related metabolites. LAC treatment counteracted stress-induced behavioral changes in dominant mice, and normalized levels of taurine, phosphocreatine, glutamine and phosphocholine in the NAc. No major accumbal metabolic changes were observed in subordinate mice.\n\nConclusionHigh social rank is confirmed as a vulnerability factor to develop chronic stress-induced depressive-like behaviors. We reveal a metabolic signature in the NAc for the antidepressant-like effects of LAC in vulnerable mice, characterized by restoration of stress-induced alterations in neuroenergetics and lipid function.

neuroscience↗

bioRxiv2019

Predictors of colony extinction vary by habitat type in social spiders

Many animal societies are susceptible to mass mortality events and collapse. Elucidating how environmental pressures determine patterns of collapse is key for our understanding of social evolution. Using the social spider Stegodyphus dumicola we investigated the environmental drivers of colony extinction along two precipitation gradients across southern Africa, using the Namib and Kalahari deserts versus wetter savanna habitats to the north and east. We deployed experimental colonies (n = 242) along two 800km transects and returned to assess colony success in the field after two months. Specifically, we noted colony extinction events after the two-month duration and collected environmental data on the correlates of those extinction events (e.g., evidence of ant attacks, # prey captured). We found that colony extinction events at desert sites were more frequently associated with attacks by predatory ants as compared to savanna sites, while colony extinctions in wetter savannas sites were more tightly associated with fungal outbreaks. Our findings support the hypothesis that environments vary in the selection pressures that they impose on social organisms, which may explain why different social phenotypes are often favored in each habitat.

ecology↗

bioRxiv2019

COLLECTIVE AGGRESSIVENESS LIMITS COLONY PERSISTENCE IN HIGH BUT NOT LOW ELEVATION SITES IN AMAZONIAN SOCIAL SPIDERS

Identifying the traits that foster group survival in contrasting environments is important for understanding local adaptation in social systems. Here we evaluate the relationship between the aggressiveness of social spider colonies and their persistence along an elevation gradient using the Amazonian spider, Anelosimus eximius. We found that colonies of A. eximius exhibit repeatable differences in their collective aggressiveness, and that colony aggressiveness is linked with persistence in a site-specific manner. Less aggressive colonies are better able to persist at high-elevation sites, which lack colony-sustaining large-bodied prey, whereas colony aggression was not related to chance of persistence at low-elevation sites. This suggests resistance to resource limitation through docility promotes colony survival at high elevations. These data reveal that the collective phenotypes that relate to colony persistence vary by site, and thus, the path of social evolution in these environments is likely to be affected.

evolutionary biology↗