Biology subjects

Qureshi, A.

Publications and source records attributed to Qureshi, A..

9 recordsLinked to original sources

bioRxiv2026

MetaResNet: Enhancing Microbiome-Based Disease Classification through Colormap Optimization and Imbalance Handling

Image-based representations of metagenomic data enable convolutional neural network (CNN) applications for microbiome disease classification. However, the impact of colormap selection on model performance remains unexplored. Current approaches arbitrarily select visualization parameters despite evidence that colormap choices can suppress minority-class features in imbalanced microbiome datasets. This study systematically evaluates colormap effects on metagenomic disease classification to establish evidence-based visualization guidelines. We developed MetaResNet, a custom CNN architecture incorporating residual blocks and attention mechanisms, to assess five colormap schemes (Jet, YlGnBu, Reds, Paired and nipy spectral) across four benchmark datasets: inflammatory bowel disease (n=110), colon cancer (n=121), women type 2 diabetes (n=96), and obesity (n=253). Class imbalance was addressed using Synthetic Minority Over-sampling Technique (SMOTE) versus class weighting strategies. Custom data augmentation preserved taxonomic abundance relationships while enhancing generalization. Performance evaluation employed F1-score, Receiver Operating Characteristic and Area Under the Curve (AUC-ROC), Matthews correlation coefficient (MCC), precision, and recall to address accuracy limitations in imbalanced scenarios. Results identified the Jet colormap coupled with SMOTE as the optimal global configuration, maximizing signal retention and achieving peak performance (AUC 1.00 in Colon). SMOTE significantly improved minority-class recall over class weighting (0.81 {+/-} 0.09 vs. 0.69 {+/-} 0.11, p = 0.003). MetaResNet achieved performance comparable to current state-of-the-art frameworks, while statistically outperforming established deep learning baselines (e.g., DeepMicro, PopPhy-CNN; p = 0.025) in discriminatory power (AUC), with peak values exceeding 0.96. These findings demonstrate that visualization efficacy is strategy-dependent, establishing MetaResNet as a robust framework for microbiome-based diagnostics that supports evidence-based visualization strategies for precision medicine.

bioinformatics↗

bioRxiv2026

DynaBiomeX: An Interpretable Dual-Strategy Deep Learning Framework for Architectural Noise Filtration in Sparse Longitudinal Microbiome Data

ObjectiveLongitudinal microbiome datasets present unique challenges due to extreme sparsity, zero-inflation, and non-stationary behavior. Conventional Recurrent Neural Networks (RNNs) struggle to distinguish structural from sampling zeros in these contexts, limiting their utility for Clinical Decision Support (CDS). MethodsWe introduce DynaBiomeX, an interpretable framework specifically developed for sparse biomedical time-series. It integrates Stacking Ensembles (Bi-LSTM, GRU) with an adapted Temporal Fusion Transformer (TFT) in a unified Screener-Sentinel workflow. The Ensembles optimize collective decision boundaries to maximize sensitivity and minimize missed cases. Concurrently, the TFT functions as a Physiological Gatekeeper, utilizing Gated Residual Networks (GRN) to actively filter stochastic noise from real biological signals. We validated this approach on a multi-modal dataset of 1,871 hematopoietic cell transplantation (HCT) patients to detect gut dysbiosis. ResultsStacking ensembles maximized discriminative performance (ROC-AUC = 0.912), effectively serving as high-sensitivity screeners. In contrast, the Adapted TFT functioned as a precision Sentinel, achieving zero false positives (Precision = 1.0) and high stability (MCC = 0.646). Crucially, the TFT demonstrated superior probabilistic reliability with a low Expected Calibration Error (ECE = 0.0085), addressing the "black-box" overconfidence typical of deep learning models. Ablation studies confirmed predictive robustness even without clinical covariates (ROC-AUC > 0.81). ConclusionDynaBiomeX couples sensitive screening with precise, calibrated validation to robustly analyze sparse longitudinal data. Validated on microbiome dysbiosis, this framework offers a scalable template for zero-inflated domains like single-cell sequencing and EHR monitoring. HighlightsO_LIDynaBiomeX integrates Stacking Ensembles and Temporal Fusion Transformers for microbiome risk stratification. C_LIO_LIGated Residual Networks distinguish stochastic sampling zeros from structural zeros in sparse ASV data. C_LIO_LIA novel Screener-Sentinel workflow resolves the trade-off between surveillance sensitivity and diagnostic precision. C_LIO_LIClinical metadata functions as a physiological gatekeeper, filtering noise to achieve zero false positives (MCC 0.646). C_LIO_LIThe framework minimizes alarm fatigue in Auto-FMT decision support by validating latent dysbiosis signals. C_LI

bioinformatics↗

bioRxiv2026

Causal Evidence for Left DLPFC Contributions to Working Memory, Attention, and Cognitive Control in Ageing: A cTBS/ERP Study

Ageing is characterised by progressive neurodegeneration and marked declines in working memory (WM), a domain critically dependent on the integrity of the dorsolateral prefrontal cortex (DLPFC). Although non-invasive brain stimulation is widely used to target this region, the causal contribution of the DLPFC to WM performance in older adults remains insufficiently understood. Here, we applied continuous theta burst stimulation (cTBS) to transiently inhibit the left DLPFC in twenty healthy older adults and assessed the behavioural and neurophysiological consequences during verbal and visuospatial N-back tasks. Behaviourally, moderate learning effects emerged exclusively in the verbal task following vertex stimulation, whereas learning after DLPFC stimulation remained inconclusive. Performance in the visuospatial task showed no reliable evidence of learning across conditions, consistent with greater interindividual variability and age-related decline in spatial WM. Neurophysiologically, ERP amplitudes were stable across conditions; however, marked latency perturbations were observed. In the verbal task, P200 and N200 latencies robustly predicted performance, with P200 latency significantly prolonged after DLPFC inhibition, indicating disrupted early attentional gating. N200 and P300 latency modulations further suggested load-dependent compensatory strategies, reflecting adaptive slowing of cognitive control processes. In contrast, visuospatial performance was not reliably predicted by P200 latency, although N200 and P300 latencies remained informative, indicating domain-specific differences in DLPFC involvement. Taken together, these findings provide causal evidence that left DLPFC disruption selectively interferes with the temporal coordination of verbal WM, emphasising the central role of processing speed and compensatory dynamics in cognitive ageing. ERP latency markers emerge as sensitive indices of subtle stimulation-induced perturbations in older adults.

neuroscience↗

bioRxiv2026

Membrane proteins retain native architecture through native ESI and soft-landing

Native MS offers a clear picture of membrane protein stoichiometry and interactions, but it lacks direct structural insights at high resolution. Here, we examine the extent to which solution-phase structure and architecture can be retained after native, soft-landing electrospray ion beam deposition (ESIBD) by interrogating several membrane-protein complexes of diverse folds and oligomeric states by cryoEM. The overall protein architectures with secondary structure motifs can be observed after gas-phase transfer, soft landing, and embedding in amorphous ice. Notably, we determined the structure of the ammonium transporter AmtB at sub-3 [A] resolution. It is nearly identical to the structure of the plunge-frozen control and even shows an extended C-terminal segment of AmtB, a dynamic region absent in the solution-phase structure. Our analysis shows that detergent adducts preserve membrane protein structure in vacuum by minimising destabilization of solvent-exposed regions and stabilization through additional polar contacts in vacuo. Molecular dynamics (MD) simulations support these results, suggesting that a monolayer shell of surfactant adducts avoids destabilization driven by unshielded polar residues and disruption of hydrogen bond networks. Overall, our findings provide a structural framework for integrating native MS with cryo-EM showing that gas-phase transfer and surfactant stabilisation preserves key architectural features and high-resolution structure of membrane proteins.

biophysics↗

bioRxiv2025

Balancing Efficiency and Compensation: Age-related Theta and Alpha Dynamics in High-load Working Memory

Working memory (WM) declines with age, yet older adults often preserve performance through compensatory mechanisms. Oscillatory dynamics in the theta and alpha bands are central to these adaptations, but their distinct roles across maintenance, manipulation, and recall remain unclear. We examined behavioural and neural differences in WM using electroencephalography (EEG) during verbal (Letter Span) and visuospatial (Corsi Block) tasks under retention and manipulation conditions. Behaviourally, older adults preserved performance in the Letter Span but showed deficits in the Corsi Test, suggesting domain-specific declines, while higher task demands reduced accuracy across both groups. At the neural level, younger adults exhibited stronger mid-frontal and temporal theta activity, supporting more efficient recruitment of WM processes and stronger performance associations. Older adults showed attenuated theta power, indicative of neural saturation, and appeared to compensate through alpha-mediated mechanisms during the delay period (maintenance / manipulation): reduced alpha suppression in the Letter Span facilitated rehearsal and updating, whereas increased upper alpha in the Corsi Test protected stored visuospatial information. These findings suggest theta oscillations underpin core WM functions, whereas alpha oscillations flexibly support compensatory strategies in ageing.

neuroscience↗

bioRxiv2025

Boosting Brainpower in Ageing: Task-Specific and Transfer Effects of tDCS-Enhanced Working Memory Training

Ageing is associated with neural alterations that impair cognitive abilities, particularly working memory (WM)--the capacity to temporarily store and manipulate information essential for daily functioning. Transcranial Direct Current Stimulation (tDCS) has shown promise in enhancing WM in older adults by modulating cortical excitability and promoting neuroplasticity. However, findings on the combined effects of tDCS and WM training remain inconsistent, particularly regarding transfer effects to non-trained WM tasks (near transfer) or other cognitive domains (far transfer). This study examined the behavioural and neural effects of tDCS on distinct WM phases (encoding, retention/manipulation, recall) in older adults. Over three weeks, participants underwent six sessions of adaptive WM training paired with tDCS targeting the left dorsolateral prefrontal cortex (DLPFC). Neural activity was assessed via EEG, focusing on theta and alpha bands. Participants were divided into tDCS and placebo groups and completed two WM tests, the Letter Span and Corsi Test, with retention and manipulation conditions. Results showed no additional behavioural or neural benefits of tDCS. However, both groups exhibited near-transfer effects that persisted one-month post-training. Neural changes varied by task: the Letter Span showed increased alpha power during manipulation, indicating enhanced maintenance of information during that phase, while the Corsi Test showed reduced left-hemisphere theta power during recall, reflecting reduced hemispheric asymmetry. Inter-individual variability likely contributed to inconsistencies and a lack of correlations between outcomes. These findings highlight the importance of personalised tDCS protocols and suggest adaptive WM training can yield enduring cognitive and neural benefits in older adults.

neuroscience↗

bioRxiv2024

Insights into the Molecular Mechanism of Pulmonary Vein Stenosis in Pediatric Patients

BackgroundPediatric pulmonary vein stenosis (PVS) is associated with substantial morbidity and mortality for the subset of patients with recurrent or progressive disease. The molecular mechanisms underlying the development and trajectory of PVS remain unclear. This study characterizes the transcriptome of clinical and phenotypic subtypes of PVS. MethodsBulk RNA sequencing analysis was performed on human pulmonary vein tissue samples obtained from surgical interventions for pediatric patients with PVS. Transcriptomic profiles were compared for primary versus post-repair PVS as well as aggressive versus non-aggressive clinical phenotypes. Principal component analysis was performed, the differential gene expression quantified, and pathway analysis conducted based on gene ontology, KEGG and Reactome. ResultsWhen comparing AggPPVS vs NonAggPPVS, differences were noted in the genes associated in extracellular matrix regulation and PIEZO1, a mechanosensitive receptor present in endothelial cells. In addition, there were notable changes in cardiac muscle contractility, calcium handling, respiratory and energy metabolism. These results point to a potential mechanism for aggressive PPVS phenotype, due to an overexpression of PIEOZ1 in response to elevated shear stress, subsequent activation of intracellular signaling pathways and leading to reduced contractility and intracellular calcium transients with cardiomyocytes. ConclusionsThe result of this study suggests that aggressive PPVS phenotype is caused by an increase in PIEZO1 expression and subsequent changes in extracellular matrix production, heart muscle contractility and changes in calcium transients within cardiomyocytes. These results provide a potential target for therapeutic invention for primary PPVS by inhibiting the activity of PIEZO1. This could potentially reduce morbidity in this patient population.

genomics↗

bioRxiv2019

Size, not temperature, drives cyclopoid copepod predation of invasive mosquito larvae

During range expansion, invasive species can experience new thermal regimes. Differences between the thermal performance of local and invasive species can alter species interactions, including predator-prey interactions. The Asian tiger mosquito, Aedes albopictus, is a known vector of several viral diseases of public health importance. It has successfully invaded many regions across the globe and currently threatens to invade regions of the UK where conditions would support seasonal activity. We assessed the functional response and predation efficiency (percentage of prey consumed) of the cyclopoid copepods Macrocyclops albidus and Megacyclops viridis from South East England, UK against newly-hatched French Ae. albopictus larvae across a relevant temperature range (15, 20, and 25{degrees}C). Predator-absent controls were included in all experiments to account for background prey mortality. We found that both M. albidus and M. viridis display type II functional response curves, and that both would therefore be suitable biocontrol agents in the event of an Ae. albopictus invasion in the UK. No significant effect of temperature on the predation interaction was detected by either type of analysis. However, the predation efficiency analysis did show differences due to predator species. The results suggest that M. viridis would be a superior predator against invasive Ae. albopictus larvae due to the larger size of this copepod species, relative to M. albidus. Our work highlights the importance of size relationships in predicting interactions between invading prey and local predators.

ecology↗

bioRxiv2019

Male competition and the evolution of mating and life history traits in experimental populations of Aedes aegypti

Aedes aegypti is an important disease vector and a major target of reproductive control efforts. We manipulated the opportunity for sexual selection in populations of Ae. aegypti by controlling the number of males competing for a single female. Populations exposed to higher levels of male competition rapidly evolved higher male competitive mating success relative to populations evolved in the absence of competition, with an evolutionary response visible after only five generations. We also detected correlated evolution in other important mating and life history traits, such as acoustic signalling, fecundity and body size. Our results indicate that there is ample segregating variation for determinants of male mating competitiveness in wild populations and that increased male mating success trades-off with other important life history traits. The mating conditions imposed on laboratory-reared mosquitoes are likely a significant determinant of male mating success in populations destined for release.

evolutionary biology↗