Browse Articles
Discover research articles across all indexed journals
Machine learning-based prediction of E. coli infection in hospitalized patients using a no-code analytical framework
Abstract Hospital-acquired infections (HAIs) remain a major global concern, contributing significantly to increased morbidity, mortality, and healthcare costs. Among the causative pathogens, Escherichia coli ( E. coli ) is one of the most frequently isolated microorganisms, particularly in urinary tract infections (UTIs), bloodstream infections, and surgical site infections. Early and accurate prediction of E. coli infection in hospitalized patients remains a significant clinical challenge, yet it has the potential to substantially improve patient outcomes. In addition, identifying patient-related risk factors can support targeted infection control strategies. This study aims to evaluate a no-code machine learning (ML) approach for early prediction of E. coli infection and to identify associated risk factors. ML techniques provide a powerful alternative by enabling the analysis of high-dimensional and heterogeneous datasets, facilitating the discovery of hidden patterns and supporting individualized risk prediction. In this study, a total of 300 clinical samples was collected as a training dataset from hospitalized patients between July 2024 and February 2025 across multiple units of Zagazig University Hospital, Sharkia, Egypt. An independent internal validation dataset of 100 samples was collected during May 2026 from the same hospital, its purpose was to evaluate model generalizability on completely unseen data. Bacterial isolates were identified using standard biochemical methods. Data analysis was performed using the Orange visual programming platform, implementing a modular ML pipeline that integrates data preprocessing, feature handling, model training, and performance evaluation within a no-code environment. The Naive Bayes model, shows potential for predicting E. coli infection in hospitalized patients. The model is intended to predict E. coli infection at the time of specimen collection, before culture results are finalized, depending on clinical data. However, further validation in larger, multi-center prospective cohorts is needed before clinical implementation.
Positional interpretation of cis-regulatory code and nucleosome organization with deep learning models
Abstract Sequence-to-function neural networks learn cis-regulatory sequence rules driving many types of genomic data. Interpreting these models to relate the sequence rules to underlying biological processes remains challenging, especially for complex genomic readouts such as MNase-seq, which maps nucleosome occupancy but is confounded by experimental bias. Here, we introduce pairwise influence by sequence attribution (PISA), which uses attribution to combinatorially decode which bases contributed to the readout at a specific genomic coordinate. PISA visualizes the effects of transcription factor motifs, detects undiscovered motifs with complex contribution patterns, and reveals experimental biases. By learning the bias for MNase-seq, PISA enables unprecedented nucleosome prediction models. These models allow the de novo discovery of nucleosome-positioning motifs and reveal the basis of Micro-C chromatin domain boundaries through systematic motif perturbations. Finally, these models allow the design of sequences with altered nucleosome configurations. These results show that PISA is a versatile tool that expands our ability to train and interpret sequence-to-function neural networks on genomics data and understand the underlying cis-regulatory code.
Ginger essential oil nanoemulsions outperform ginger extract nanoemulsions in extending the shelf life of tilapia fillets
Quantum interference in a twisted high-Tc SQUID senses emergent interfacial order
A dual-strain compartmental model for Mycoplasma pneumoniae transmission dynamics with vulnerability stratification and optimal control strategies
Octahedral-motif-guided design of optoelectronic semiconductors via interpretable machine learning
Fabrication of hybrid PCL-Chitosan/PVA nanofibers by hybrid electrospinning for Local melanoma skin cancer therapy
The burden of antimicrobial-resistant bacterial infections: a causal perspective
Granular soil densification using expansive resins
A likelihood-based method for identifying replay from spike sequences
An energy-efficient clustering and routing in WSN using density based adaptive soft clustering with adaptive lotus effect optimization and IGGO
A large-scale multi-ancestry mitochondrial variant association analysis for cardiometabolic traits
Abstract Studies linking mitochondrial DNA (mtDNA) with complex traits are often limited by small sample sizes or focused on specific phenotypes in clinically selected cohorts. Here, we use data from >600,000 participants in the Million Veteran Program (MVP) to perform a multi-ancestry analysis of mitochondrial DNA (mtDNA) variation and cardiometabolic phenotypes across European (EUR), African (AFR), Admixed American (AMR), and East Asian (EAS) populations. After validating 248 mtDNA loci, we identify 10 ancestry-stratified single-variant associations (8 EUR, 2 AFR) and 23 additional signals in sex- and type 2 diabetes (T2D)–stratified analyses. Four variants tagging haplogroup J, D-loop MT228G > A, MT-ND3 MT10398A > G (p.Thr114Ala), MT-ND5 MT13708G > A (p.Ala458Thr), and MT-CYB MT14798T > C (p.Phe18Leu), are associated with hypothyroidism in EUR and replicated in UK Biobank (UKBB) with concordant effects. In EUR females, MT-RNR1 MT1555A > G increases the risk of carditis and heart failure phenotypes, supporting prior reports of maternally inherited cardiomyopathy. Gene-based rare-variant tests (minor allele frequency ≤2%) yield 26 associations (12 EUR, 10 AFR, 2 AMR, 2 EAS), including mitochondrial tRNA burdens linked to primary cardiomyopathy (females) and exophthalmos (males). Twenty-three of the 33 single-variant signals map to the endocrine/metabolic category, indicating significant enrichment (Fisher’s exact P = 0.003). These results define ancestry- and context-specific contributions of mtDNA to cardiometabolic disease, with a notable concentration in endocrine traits, and provide a framework for mtDNA analysis across diverse biobank cohorts.
Valorization of Moringa oleifera pericarp via semi-synthetic sugar-based enone derivatives with anticancer potential: phytochemical isolation, cytotoxic evaluation, and dual EGFR/CAIX targeting
Abstract Moringa oleifera pericarp, an underutilized agro-waste, was investigated as a potential source of anticancer agents. In this study, bioassay-guided fractionation of the pericarp extract led to the identification of the ethyl acetate fraction as the most active fraction, showing notable antioxidant, antimicrobial, and cytotoxic activities. Subsequent purification afforded three compounds: 4-( α -L-rhamnopyranosyloxy)-benzaldehyde ( M 1 ), 4-( α -L-rhamnopyranosyl) benzyl alcohol ( M 2 ), and 4-(hydroxymethyl) phenol-1- O - β -D-glucopyranosyl-(1''→3')- O - α -L-rhamnopyranoside ( M 3 ). The major constituent, M 2 , was semi-synthetically modified by Steglich esterification with cinnamic acid and crotonic acid to yield two sugar-based enone derivatives, S1 M2 and S2 M2 . Structural elucidation was performed using IR, MS, and NMR spectroscopy. The cytotoxic activity of the isolated and semi-synthesized compounds was evaluated against HepG2 and HCT116 cancer cell lines, and selectivity was assessed using normal WI-38 fibroblasts. Among the tested compounds, S2 M2 exhibited the most potent cytotoxicity, with IC 50 values of 5.97 ± 0.19 µM and 11.52 ± 0.37 µM against HepG2 and HCT116 cells, respectively, together with favorable selectivity. In addition, S2 M2 showed strong inhibitory activity against epidermal growth factor receptor tyrosine kinase (EGFR-TK) and carbonic anhydrase IX (CAIX), with IC 50 values of 0.40 ± 0.008 µM and 0.27 ± 0.01 µM, respectively. Molecular docking and 100-ns molecular dynamics simulations supported the stable binding of S2 M2 within the active sites of EGFR-TK and CAIX, and MM-GBSA calculations confirmed its favorable binding free energy. Structure–activity relationship analysis suggested that the α , β -unsaturated carbonyl moiety significantly contributed to the observed anticancer activity. These findings highlight M. oleifera pericarp as a promising source of bioactive glycosylated phenolic scaffolds and identify S2 M2 as a potential dual-target anticancer lead for further development.
Sub-parts-per-billion CO2 Detection based on Dissipative Whispering Gallery Mode Microcavity Sensor
Abstract Whispering gallery mode microcavities provide strong light–matter interactions owing to their ultrahigh optical confinement, but the small gas refractive index change limits their ability to sense trace gases. Here we show that gas absorption can be detected using a dissipative sensing mechanism in a non-functionalized whispering gallery mode microcavity. Instead of tracking resonance frequency shifts used in conventional dispersive sensing, our method converts optical absorption into variations in resonance depth through thermally induced dissipation. Quantitative carbon dioxide detection was achieved over a concentration range of 1.5 to 400 parts per million with a correlation coefficients exceeding 0.99. The sensor reached a detection limit of 168 parts per trillion at an integration time of 400 seconds and an accuracy of approximately 0.4%. Continuous monitoring further demonstrated stable operation under ambient conditions. These results establish dissipative microcavity sensing as a promising approach for compact, low-cost, and highly sensitive trace gas detection.