Machine Learning Prediction of Intestinal α-Glucosidase Inhibitors Using a Diverse Set of Ligands: A Drug Repurposing Effort with DrugBank Database Screening DOI Creative Commons
Adeshina I. Odugbemi, Clement N. Nyirenda, Alan Christoffels

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: April 18, 2024

Abstract The global rise in diabetes mellitus (DM) poses a significant health challenge, necessitating effective therapeutic interventions. α-Glucosidase inhibitors play crucial role managing postprandial hyperglycemia and reducing the risk of complications Type 2 DM. Quantitative Structure-Activity Relationship (QSAR) modeling is critical computational drug discovery. However, many QSAR studies on α-glucosidase often rely limited compound series statistical methods, restricting their applicability across wide chemical space. Integrating machine learning (ML) into offers promising avenue for discovering novel compounds by handling complex information from diverse sets. Our study aimed to develop robust predictive models using dataset 1082 with known activity against intestinal (maltase-glucoamylase). After thorough data preparation, we employed 626 train ML models, generating different training three distinct molecular representations: 2D-descriptors, 3D-descriptors, Extended-connectivity-fingerprint (ECFP4). These trained random forest support vector algorithms, underwent rigorous evaluation established metrics. Subsequently, best-performing model was utilized screen Drugbank database, identifying potential inhibitor drugs. Drug repurposing, an expedited strategy new uses existing drugs, holds immense this regard. Molecular docking dynamics simulations further corroborated our predictions. results indicate that 2D descriptors ECFP4 representations outperform 3D descriptors. Furthermore, candidates identified DrugBank screening exhibited binding interactions α-glucosidase, corroborating predictions supporting repurposing.

Language: Английский

Human milk oligosaccharide secretion dynamics during breastfeeding and its antimicrobial role: A systematic review DOI
Mohammed Al‐Beltagi

World Journal of Clinical Pediatrics, Journal Year: 2025, Volume and Issue: 14(2)

Published: March 18, 2025

BACKGROUND Human milk oligosaccharides (HMOs) are bioactive components of breast with diverse health benefits, including shaping the gut microbiota, modulating immune system, and protecting against infections. HMOs exhibit dynamic secretion patterns during lactation, influenced by maternal genetics environmental factors. Their direct indirect antimicrobial properties have garnered significant research interest. However, a comprehensive understanding dynamics their correlation efficacy remains underexplored. AIM To synthesize current evidence on lactation evaluate roles bacterial, viral, protozoal pathogens. METHODS A systematic search PubMed, Scopus, Web Science, Cochrane Library focused studies investigating natural synthetic HMOs, dynamics, properties. Studies involving human, animal, in vitro models were included. Data HMO composition, temporal patterns, mechanisms action extracted. Quality assessment was performed using validated tools appropriate for study design. RESULTS total 44 included, encompassing research. exhibited 2′-fucosyllactose (2′-FL) lacto-N-tetraose peaking early declining over time, while 3-fucosyllactose (3-FL) increased later stages. demonstrated through pathogen adhesion inhibition, biofilm disruption, enzymatic activity impairment. Synthetic bioengineered 2′-FL 3-FL, structurally functionally comparable to effectively inhibiting pathogens such as Pseudomonas aeruginosa , Escherichia coli Campylobacter jejuni . Additionally, synergistic effects antibiotics, enhancing resistant CONCLUSION vital defense, supporting infant targeting various Both hold potential therapeutic applications, particularly nutrition adjuncts antibiotics. Further research, clinical trials, is essential address gaps knowledge, validate findings, explore broader applicability improving neonatal health.

Language: Английский

Citations

0
Genetic Diversity in the Fusion Gene of Respiratory Syncytial Virus (RSV) Isolated From Iraqi Patients: A First Report DOI Creative Commons
Mohammed Hussein Wali, Hassan M. Naif, Nur Arzuar Abdul Rahim

et al.

Advances in Virology, Journal Year: 2025, Volume and Issue: 2025(1)

Published: Jan. 1, 2025

Molecular evaluation of the respiratory syncytial virus (RSV) genome is one common strategies applied to understand viral pathogenicity and control its spreading. In this study, we carried out molecular on targeted fusion (F) gene region in RSV‐positive samples Iraqi patients during autumn winter 2022/2023. One hundred fifty with lower tract infections were screened for RSV using reverse transcription‐quantitative polymerase chain reaction (RT‐qPCR). Sanger sequencing was performed targeting 1061 nucleotides (from nucleotide 6168 7228 within genome) 1000 6122 7121 F RSV‐A RSV‐B, respectively. The results showed some changes gene, which grouped distinct clade, closely related isolates from Austria, Argentine, Finland, France through phylogenetic analysis. silico protein modeling SWISS‐MODEL I‐TASSER web tools based nonsynonymous amino acid sequence good‐predicted models that can be utilized antiviral screening. summary, identified variations could influence vaccine development as primary target major antigen RSV. surveillance data local are also essential studying new genomic enable prediction potential agents.

Language: Английский

Citations

0
Human milk oligosaccharide profiles remain unaffected by maternal pre-pregnancy body mass index in an observational study DOI Creative Commons
Julie Astono, Yu-Ping Huang, Ulrik Kræmer Sundekilde

et al.

Frontiers in Nutrition, Journal Year: 2024, Volume and Issue: 11

Published: Oct. 16, 2024

Human milk oligosaccharides (HMOs) are important carbohydrates in human that infants cannot digest, acting as prebiotics linked to infant health. The risk of childhood obesity increases with maternal obesity, potentially mediated through the gut microbiota affected by available HMOs. Studies on whether affects HMO abundance, yield conflicting results. This study aimed investigate profile and its association measured pre-pregnancy body mass index (BMI) anthropometrics. results were discussed context existing literature. 90 samples collected at 3 months postpartum from mothers three BMI-groups: 32 normal weight (BMI: 18.5–24.99 kg/m 2 ), 34 overweight 25–30 24 obese (BMI > 30 ). analyzed using nano liquid chromatography chip quadrupole time-of-flight spectrometry yielding 51 structures isomers. Their peak areas integrated normalized determine relative abundances. Univariate multivariate analysis showed associations between abundance donors’ secretor status specific anthropometric variables, but not BMI. does support hypothesis influences highlights importance reporting despite absence significant correlations.

Language: Английский

Citations

1
Machine Learning Prediction of Intestinal α-Glucosidase Inhibitors Using a Diverse Set of Ligands: A Drug Repurposing Effort with DrugBank Database Screening DOI Creative Commons
Adeshina I. Odugbemi, Clement N. Nyirenda, Alan Christoffels

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: April 18, 2024

Abstract The global rise in diabetes mellitus (DM) poses a significant health challenge, necessitating effective therapeutic interventions. α-Glucosidase inhibitors play crucial role managing postprandial hyperglycemia and reducing the risk of complications Type 2 DM. Quantitative Structure-Activity Relationship (QSAR) modeling is critical computational drug discovery. However, many QSAR studies on α-glucosidase often rely limited compound series statistical methods, restricting their applicability across wide chemical space. Integrating machine learning (ML) into offers promising avenue for discovering novel compounds by handling complex information from diverse sets. Our study aimed to develop robust predictive models using dataset 1082 with known activity against intestinal (maltase-glucoamylase). After thorough data preparation, we employed 626 train ML models, generating different training three distinct molecular representations: 2D-descriptors, 3D-descriptors, Extended-connectivity-fingerprint (ECFP4). These trained random forest support vector algorithms, underwent rigorous evaluation established metrics. Subsequently, best-performing model was utilized screen Drugbank database, identifying potential inhibitor drugs. Drug repurposing, an expedited strategy new uses existing drugs, holds immense this regard. Molecular docking dynamics simulations further corroborated our predictions. results indicate that 2D descriptors ECFP4 representations outperform 3D descriptors. Furthermore, candidates identified DrugBank screening exhibited binding interactions α-glucosidase, corroborating predictions supporting repurposing.

Language: Английский

Citations

0