Papua New Guinea University of Technology

The rising prevalence of diabetes necessitates accurate, noninvasive methods for early detection and personalized management. Breathomics, the investigation of exhaled volatile organic compounds (VOCs), offers a promising approach for evaluating glycemic control, oxidative stress, and metabolic alterations associated with diabetes. Specific VOCs, such as acetone, isoprene, and ethanol, reflect glucose fluctuations, lipid oxidation, and insulin sensitivity, linking metabolic changes to real-time physiological states. Recent analytical advancements in gas chromatography-mass spectrometry, proton transfer reaction-mass spectrometry, and electronic nose technologies have enhanced the reliability and clinical applicability of breath-based assessments. The investigations focus on integrating VOC-based metabolic signatures with lifestyle and behavioral factors to establish a precision, noninvasive framework for diabetes management. This review also discusses key challenges that remain in standardizing breath analysis techniques, interpreting complex breath data, and implementing breath-based technology into personalized diabetes management. Breathomics provides a novel pathway for dynamic real-time assessment of metabolic health, bridging physiological biomarkers with lifestyle responses to enable individualized diabetes management.

Diabetes mellitus is a chronic metabolic disorder that affects millions of people globally. Among three types of diabetes, Type 2 diabetes mellitus (T2DM) is a rapidly growing global health challenge. Despite available modern antidiabetic drugs, patients still struggle with side effects and treatment failure, as an alternative to this, there is a crucial requirement to develop a potential and traditional plant-based medicine which could be a safer sources and multi-target therapies to treat chronic disease like diabetes. Solanum lasiocarpum (S. lasiocarpum) is a sour fruit-vegetable being widely used in Southeast Asia as both food and traditional medicine, including for the management of diabetes. However, its active components and antidiabetic mechanisms have not been systematically explored. In this study, we combined metabolomics, proteomics, and transcriptomics to investigate the bioactive pathways and potential molecular targets of S. lasiocarpum. Untargeted UHPLC–QTOF–MS profiling identified 45 candidate bioactive compounds with good predicted gastrointestinal absorption, and the network pharmacology analysis linked these compounds to 43 diabetes-related human targets. Protein–protein interaction analysis highlighted several core nodes, including TNF, PPARG, IL6, AKT1, and STAT3, and functional enrichment suggested roles in hormone regulation, inflammation, glucose and lipid metabolism, and vascular function. De novo transcriptome assembly and data-independent acquisition-based proteomics of mature S. lasiocarpum fruit showed that central carbon metabolism is highly active and that the shikimate, phenylpropanoid, and flavonoid pathways are strongly expressed at both gene and protein levels. Key enzymes such as EPSPS, PAL, C4H, 4CL, CHS, CHI, F3H, and FLS formed a coherent biosynthetic network supporting sustained production of phenolic and flavonoid metabolites. Integrating these omics layers with target prediction suggests that S. lasiocarpum may exert antidiabetic effects by modulating a TNF–PPARG axis, reducing pro-inflammatory signaling while supporting insulin-sensitizing pathways. Overall, these results support the traditional use of S. lasiocarpum and provide a multi-omics resource to prioritise candidate metabolites, enzymes and targets for follow-up studies. As the pathway links were inferred computationally, the proposed TNF–PPARG-centred mechanism should be regarded as hypothesis-generating and will require validation in experimental models and, ultimately, well-designed human intervention trials.

Malachite green–contaminated wastewater discharged without proper treatment poses serious risks to human health and the environment. Therefore, this study evaluated the adsorptive removal of malachite green from aqueous solution using activated carbon derived from anchote peel. The adsorbent was prepared via chemical activation with phosphoric acid followed by thermal activation in the absence of oxidizing agents. Adsorption optimization was performed using the Box–Behnken design of response surface methodology with four variables: pH (3, 6, 9), initial dye concentration (50, 100, 150 mg/L), adsorbent dosage (0.5, 1.0, 1.5 g/100 mL), and contact time (30, 60, 90 min), resulting in 30 experimental runs. The maximum dye removal efficiency of 99.07% was achieved at pH 9, 50 mg/L initial concentration, 1.5 g/100 mL dosage, and 90 min contact time. ANOVA confirmed strong agreement between adjusted and predicted values (R² = 0.9935 and 0.9917) based on the quadratic regression model. These findings demonstrate that anchote peel–based activated carbon is an efficient adsorbent for malachite green removal, and its optimization using the Box–Behnken design provides a promising approach for large-scale treatment of dye-laden industrial wastewater

A study was conducted on the engineering properties of anchote (Coccinia abyssinica) starch-based biodegradable films developed using plasticizers (glycerols) for food packaging. Compared to films derived from other common starch sources, anchote starch films exhibited superior mechanical and optical properties, making them a promising alternative for sustainable packaging. The experiment was planned with two factors, glycerol concentration and starch concentration, and consisted of nine combinations of treatments executed in a factorial experiment. The plastic's physical, mechanical, and optical properties were evaluated with the glycerol concentrations of 20, 25, and 30% and the starch concentrations of 3, 5, and 7 g/100 mL. The Anchote starch plastic films demonstrated 9.39–11.89% moisture content, water solubility of 71.31–74.56%, and water absorption of 23.43–33.58%. The films also showed tensile strength values of 10.86–12.20 MPa, elongation at break between 27.14% and 30.03%, and a Young’s modulus ranging from 12.08 to 12.84 MPa. The thickness varied from 0.41 to 0.49 mm, transparency ranged from 79.39% to 87.41%, and lightness was between 81.48% and 83.78%. The biodegradation rate ranged from 92.49% to 98.62%. These properties were significantly influenced by both glycerol and starch concentrations (P < 0.05). Additionally, density (0.97 to 1.05 g/cm 3 ), color values a*(-0.33 to -0.37), b*(3.10 to 3.22), and ΔE (13.83 to 15.46), and the white index (82.33 to 83.99) were significantly (P < 0.05) affected by the glycerol concentration only. Generally, the study verified that the biodegradable films' properties were affected by the glycerol concentration and starch concentration. The mechanical and optical properties of the same film showed that the film was suitable for food packaging.

This study aimed to identify leguminous green manure species with strong adaptability, high nutrient accumulation, and effective weed suppression for subsistence farming in Papua New Guinea. It also examined the optimal termination age of 45 and 65 days after planting (DAP) for lablab, velvet bean, calapo, centro, and puero. The agronomic performance was compared to natural vegetation in a field experiment laid out in a Randomized Complete Block Design. Velvet bean produced the highest aerial dry matter (11.2 t ha-1), leading to significantly greater accumulation of the primary nutrients nitrogen (374 kg ha-1), phosphorus (61 kg ha-1), and potassium (119 kg ha-1) within 65 days. The green manure crops evaluated derived 21.8%–87.7% of their N from atmospheric fixation. Within 28 DAP, velvet bean and lablab suppressed weed growth by 87% and 53%, respectively. Termination age significantly affected the nutrient accumulation in velvet bean, suggesting 65 DAP could be ideal for soil incorporation, whereas lablab can be terminated at 45 DAP. Our findings suggest that integrating velvet bean and lablab in improved fallow systems could enhance soil fertility and weed management in humid tropical regions.

Agile Project Management (APM) has become fundamental to providing flexible, efficient, and iterative project solutions. However, as the projects evolve in complexity, managing them with the use of the traditional Agile methods can be very unproductive this is due to the following vices; loose planning, improper resource allocation, and very late risk identification among others. Applying artificial intelligence (AI) in the context of Agile practices has brought about improvements to innovation by breaking barriers and expanding the advantages of Agile practices. This paper aims to discuss the future vision of APM that real-time AI tools enhance through down-to-earth AI applications during sprint planning, resource assignments, and risk evaluation by way of predictive and prescriptive analytics. The use of AI promotes innovation through the creation of digital environments for automating routine processes, for supporting decisions based on research on the implications of an AI-supported process in detail, and for enhancing self-organized models of communication within working groups. These include; reducing bias in taking estimations, increasing scalability for major projects, and adaptability in case of changing needs. Moreover, AI extends the value of Agile by cutting down project time, raising the cost-tovalue ratio, and satisfying stakeholders to the greatest extent possible. Features compare AI-based APM and conventional APM and the results show that the former outperforms the latter in all aspects including Accuracy, adaptability, and transparency. Kyriakou presents real-world examples from IT and Startup industries on how AI enhances Agile processes and the resulting improvement of the project. Nonetheless, such developments present implementing hurdles such as privacy issues, system integration, and organizational enmity. Regarding these, the study gives practical implications: proposing the reference model for AI implementation, how to incorporate AI into the Agile approach, increasing team preparedness, and engaging stakeholders.

Federated Learning (FL) which eliminates the centralized data storage requirement by facilitating model training on diverse edge devices is now a promising paradigm for decentralized machine learning (ML). Applications involving privacy-preserving Artificial Intelligence (AI), including wearable technology, IoT networks, and smart healthcare appliances, can particularly benefit from this solution in embedded systems. By using on-device local data from devices such as sensors, embedded controllers, and smartphones, FL keeps confidential information local, minimizing the data transfer cost and privacy risks. Potentiality, challenges, and key applications of FL integration with embedded systems are addressed in this paper. Device-to-device efficient communication, model updating, and trade-offs between model accuracy and computational resource limitations are some of the issues addressed. Also addressed in the paper are model aggregation, federated optimization methods, and their usage in edge-based AI in real-life applications. Problems with security, system reliability, and heterogeneous data in federated environments are also discussed in the paper. The extensive use of FL in embedded systems is one of the important developments in edge AI solution designing that is more scalable, secure, and privacy-conscious.