Biology subjects

Tariq, H.

Publications and source records attributed to Tariq, H..

7 recordsLinked to original sources

bioRxiv2026

Multi-Scale Contextual Attention for Robust Crop and Pest Image Classification

Image-based crop and pest recognition is considered useful for reducing the delay and cost of manual field scouting, therefore supporting timely intervention in precision-agriculture workflows. However, the real field imagery remains challenging due to the cluttered backgrounds, occlusions, illumination changes, and strong scale variation that are frequently observed across crops. The symptoms are often small or low-contrast, and pests may be partially hidden, which reduces the reliability when the setting is outside controlled environments. A unified multi-class crop-pest/condition recognition framework is presented, where a ResNet-50 backbone is utilized and enhanced with a Multi-Scale Contextual Attention (MSCA) module. The novelty is mainly considered to be achieved through the integration of explicit multi-scale contextual aggregation with lightweight joint channel and spatial attention by means of residual fusion, while the empirical evaluation was kept controlled under a fixed and reproducible protocol. A curated dataset of 21,404 field-style images covering 15 crop and pest/condition classes was compiled, and a leakage-aware fixed split with a held-out test set was adopted to support reproducibility. Augmentation was applied only to the training subset to improve robustness, although the validation data was not augmented in the same manner. On the held-out test set, balanced performance was achieved by the proposed approach, with about 0.93 accuracy and a macro-F1 score close to 0.94 being obtained, while established baselines such as EfficientNet, Vision Transformer, and attention-based CNN models were outperformed under identical evaluation settings. Controlled ablations were used to isolate the contribution of MSCA and augmentation under the same training configuration. These results indicate that lightweight multi-scale contextual attention is effective for crop and pest recognition under realistic field conditions, although some visually similar classes remained difficult.

plant biology↗

bioRxiv2026

Sphingosine-1-Phosphate Receptor 1 regulates competition dependent astrocyte morphogenesis and tiling in murine cortex.

Astrocytes are highly abundant in the mammalian brain and coordinate with neurons and other glial cells to regulate neural circuit structure, function, and blood brain barrier integrity among many to maintain proper brain homeostasis. Astrocytes perform most of these functions owing to their highly complex morphologies and hundreds of thousands of fine processes that are important in contacting neuronal synapses and other glial cells. In fact, the morphological complexity of astrocytes is regulated by the presence and activity of neurons and helps establish astrocyte territory/tiling in a non-overlapping pattern; however, the mechanisms of astrocyte tiling are not well characterized. Using a human astrocyte-mouse neuron coculture system, we previously showed that sphingosine-1-phosphate receptor 1 (S1PR1) regulates astrocyte morphogenesis in a neuronal contact dependent manner. In this study, we find that S1PR1, in vivo, regulates astrocyte morphogenesis in a cortical layer specific manner. Using astrocyte-specific S1PR1 knock out mouse models and adenoassociated viral labeling methods, we show that S1PR1 is crucial in establishing competition driven astrocyte tiling and morphogenesis in the developing brain. Furthermore, we show that JAK-STAT3 signaling regulates neuronal contact induced expression of S1PR1 in cocultured astrocytes. These studies therefore uncover a lipid signaling receptor as a major regulator of astrocyte morphogenesis and tiling in murine cortical layers.

neuroscience↗

bioRxiv2026

In Silico Identification of Aminoadipate Semialdehyde Synthase (AASS) as a Novel Prognostic Biomarker in Triple-Negative Breast Cancer

Triple-negative breast cancer (TNBC) is an aggressive subtype that lacks effective targeted therapies. This study aimed to identify robust prognostic biomarkers by integrating network biology with machine learning (ML) approaches. TNBC expression cohorts were analysed to identify differentially expressed genes (DEGs) and crucial gene clusters via limma and Weighted Gene Co-expression Network Analysis (WGCNA). In results, 579 DEGs were identified, and network analysis revealed two TNBC-associated modules. Overlapping determined 208 genes enriched in cell-cycle and mitotic-regulation pathways. To identify candidate biomarkers, protein-protein interaction (PPI) networks and ML feature selection techniques, including support vector machine with recursive feature elimination module (SVM-RFE) and least absolute shrinkage and selection operator (LASSO) regression, were performed. The Kaplan-Meier (KM) analysis revealed AASS and CCNA2 were favourable prognostic markers, whereas CXCL8, SPP1, and CCNB1 were poor prognostic markers. Multi-level validation and immune-subtype analysis were carried out, revealing AASS as a novel TNBC-associated metabolic tumour suppressor.

cancer biology↗

bioRxiv2025

Adipocyte IL-22RA1 signaling promotes structural and functional remodeling of white adipose tissue following acute intestinal damage.

Interleukin (IL)-22 has been shown to play an important role in intestinal host defense and ameliorating high-fat diet (HFD)-induced metabolic disorders primarily through signaling in intestinal epithelial cells. Adipocytes have emerged as key immune-metabolic regulators that influence intestinal inflammation in inflammatory bowel disease (IBD). However, the role of IL-22RA1 signaling in adipocytes has not been explored. In the present study, we examined the role of IL-22RA1 signaling in adipocytes in response to dextran sulfate sodium (DSS)-mediated gut inflammation. To do this, we subjected adipocyte specific Il22ra1 knockout mice (Il22ra1Adipo) to intestinal inflammation under normal chow and HFD conditions. Compared to littermate controls, Il22ra1Adipo mice displayed significant weight loss on day 9 of DSS treatment and showed altered expression of immune response- and lipid metabolism-related genes in white adipose tissue (WAT) under a normal chow diet. Notably, when mice were primed with HFD prior to DSS-induced intestinal injury, WAT from Il22ra1Adipo mice showed a significant reduction in Fabp4 expression and a marked increase in the proliferation marker Ki67. These findings indicate that loss of IL-22RA1 signaling in adipocytes disrupts adipocyte differentiation and lipid metabolism, leading to increased proliferation of preadipocytes or stromal cells without proper maturation. Importantly, this altered adipose tissue response occurred despite similar levels of colonic inflammation between knockout and control mice, suggesting a critical role for adipocyte IL-22RA1 signaling in maintaining metabolic and inflammatory homeostasis in WAT during combined metabolic and intestinal inflammatory stress.

pathology↗

bioRxiv2025

Comparative genomics of the lipid droplet-associated protein Seipin across eukaryotic diversity illuminates an ancient origin and conserved structural diversity.

Lipid droplets (LDs) are ubiquitous across living organisms. Characterised in eukaryotes by their lipid monolayer and essential for lipid storage and metabolism in animals, plants, and yeast, little is known about the evolutionary diversity of extant LDs across the eukaryotic tree of life. LDs facilitate an impressive variety of cellular functions outside of just lipid storage; in Metazoa, these include stress responses, cellular signaling, and membrane remodeling. Likewise, the distribution of these functions across eukaryotic diversity is unknown. We have examined the evolutionary trajectory of seipin, a protein associated with LD biogenesis, across eukaryotic diversity. We have identified a pan-eukaryotic distribution of seipin, with an evolutionary pattern that indicates presence in the Last Eukaryotic Common Ancestor. Ancient conservation of multiple variants of seipin suggests that seipin may multiple conserved functional roles within the cell. Finally, we identify a lack of sequence homology between BSCL2/seipin in Homo sapiens and the sei-1 gene previously identified as a functional homologue of BSCL2 in Saccharomyces cerevisiae. Though our results suggest that sei-1 may be a highly divergent homologue of BSCL2 / seipin rather than an unrelated protein, the divergence suggests that researchers may need to be cautious applying results obtained in Saccharomyces spp.. Visual abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=145 SRC="FIGDIR/small/683490v1_ufig1.gif" ALT="Figure 1"> View larger version (41K): org.highwire.dtl.DTLVardef@1d380c5org.highwire.dtl.DTLVardef@ab2432org.highwire.dtl.DTLVardef@1d56cb5org.highwire.dtl.DTLVardef@d3fcbc_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2025

From Metabolism to Malignancy: Profiling Diabetes-Related Genes in Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) and diabetes mellitus both affect the liver, a key metabolic organ. This study uses bioinformatics to explore genetic links between Type 1 Diabetes (T1D) and HCC prognosis. Eleven HCC gene expression datasets from GEO and one TCGA-LIHC dataset were analyzed. Gene Set Enrichment Analysis (GSEA) identified up- and downregulated genes after data normalization in R. Four datasets (GSE64041, GSE78737, GSE107170, TCGA) revealed increased expression of T1D-related genes. Ten genes, including HLA-DOB and HLA-DPB1, were consistently upregulated. Statistical analysis (Kruskal-Wallis and Mann-Whitney U tests) showed these two genes were significantly associated with tumor grade and T-stage, with p-values ranging from 0.008 to 0.019. Co-expression analysis with 96 literature-curated T1D genes identified 23 related genes. Survival analysis using Kaplan-Meier curves highlighted five genes (IL7R, CD69, CCR5, RUNX3, PRF1), with CD69 showing strong associations with T-stage and disease-free survival. PRF1, RUNX3, and CCR5 were also linked to survival outcomes. Seven diabetes-related GEO datasets were used for validation. GSEA showed T1D gene enrichment in two datasets (GSE228267, GSE232310), with HLA-DOB significantly expressed in GSE228267 (p = 0.004). These findings suggest that HLA-DOB, HLA-DPB1, and three other T1D-related genes may serve as potential biomarkers for understanding the genetic connection between T1D and HCC. Though computational, this study lays the groundwork for future experimental validation.

bioinformatics↗

bioRxiv2024

Nasal microbionts differentially colonize and elicit cytokines in human nasal epithelial organoids

Nasal colonization by Staphylococcus aureus or Streptococcus pneumoniae is associated with an increased risk of infection by these pathobionts, whereas nasal colonization by Dolosigranulum species is associated with health. Human nasal epithelial organoids (HNOs) physiologically recapitulate human nasal respiratory epithelium with a robust mucociliary blanket. We reproducibly monocolonized HNOs with these three bacteria for up to 48 hours with varying kinetics across species. HNOs tolerated bacterial monocolonization with localization of bacteria to the mucus layer and with minimal cytotoxicity compared to uncolonized HNOs. Human nasal epithelium exhibited both species-specific and general cytokine responses, without induction of type I interferons, consistent with colonization rather than infection. Only live S. aureus colonization robustly induced IL-1 family cytokines, suggestive of inflammasome signaling. D. pigrum and live S. aureus decreased CXCL10, whereas S. pneumoniae increased CXCL11, chemokines involved in antimicrobial responses to both viruses and bacteria. Overall, HNOs are a compelling model system to reveal host-microbe dynamics at the human nasal mucosa. IMPORTANCEHuman nasal microbiota often includes highly pathogenic members, many of which are antimicrobial resistance threats, e.g., methicillin-resistant Staphylococcus aureus and antibiotic-resistant Streptococcus pneumoniae. Preventing colonization by nasal pathobionts decreases infections and transmission. In contrast, nasal microbiome studies identify candidate beneficial bacteria that might resist pathobiont colonization, e.g., Dolosigranulum pigrum. Discovering how these microbionts colonize the human nasal passages and means to reduce pathobiont colonization is limited by previous models. This creates an urgent need for human-based models that exemplify bacterial nasal colonization. We addressed this need by developing human nasal epithelial organoids (HNOs) as a new model system of bacterial nasal colonization. HNOs accurately represent the mucosal surface of the human nasal passages enabling exploration of bacterial-epithelial interactions, which is crucial since the epithelium instigates the initial innate immune response to bacteria. Here, we identified differential epithelial cytokine responses to these three bacteria setting the stage for future research.

microbiology↗