Welcome to Mbeya University of Science and Technology Repository
Mbeya University of Science and Technology Repository (MUST Repository) is an open-access digital platform dedicated to the collection, preservation, management, and dissemination of the University’s official scholarly and institutional resources. Authorized by the University and aligned with national academic and research objectives, the repository ensures that valuable institutional knowledge remains securely preserved and easily accessible to researchers, students, staff, and the wider public.
Communities in MUST Repository
Select a community to browse its collections.
Recent Submissions
Distribution patterns and conservation status of anurans in Kanga forest reserve, Eastern Arc Mountains, Tanzania
(Universuty of LATVIA, 2026-06-22) Mugini, Geofrey; Ojija, Fredrick; Ojija, Fredrick
Understanding the ecological drivers of anuran assemblages is essential for biodiversity conservation in tropical montane ecosystems. This study assessed anuran diversity, distribution patterns, conservation status, and environmental determinants of species in Kanga Forest Reserve (KFR), an important component of the Eastern Arc Mountain Forests of Tanzania. Standardized surveys were conducted across four elevational categories and major habitat types to evaluate how habitat structure and environmental conditions influence anuran communities. A total of 455 individuals representing 22 species, 12 genera, and seven families were recorded. Species abundances were highest in lowland and sub-montane forests, whereas farmland and montane forests had fewer individuals. Similarly, diversity was highest in lowlands (H’ = 2.45) and lowest in farmland (H’ = 1.80), indicating anthropogenic impacts. Forest-restricted taxa, including Callulina kanga and Probreviceps cf. durirostris, were largely confined to structurally intact forest habitats, whereas species such as Chiromantis rufescens and Ptychadena anchietae exhibited broader ecological tolerances. Integrated analyses revealed that anuran distributions were influenced primarily by hydrological conditions and forest structural complexity. Water bodies, canopy cover, tree density, leaf litter, and mature trees emerged as the strongest predictor of species occurrence, closely associated with species persistence and community composition. The results therefore demonstrate that habitat complexity and moisture availability function as key ecological filters shaping community assembly within KFR. Conservation assessment revealed species spanning multiple IUCN Red List categories, including Critically Endangered, Endangered, Vulnerable, and Near Threatened taxa, highlighting the reserve’s importance as a refuge for threatened anurans. This study provides a robust baseline for long-term monitoring and delivers testable ecological evidence to guide conservation planning, habitat restoration, and adaptive management within KFR and the broader Eastern Arc Mountain ecosystem.
An explainable quantum neural network framework for reservoir permeability prediction from conventional well logs
(Elsevier, 2026-08-30) Alvin K. Mulashani; Christopher N. Mkono; Melckzedeck M. Mgimba
Accurate prediction of reservoir permeability is essential for reservoir characterization, production forecasting, and hydrocarbon recovery optimization. Conventional empirical and machine learning approaches often struggle to capture the complex nonlinear relationships between petrophysical variables while providing limited model interpretability. This study presents an explainable permeability prediction framework based on a variational Quantum Neural Network (QNN) simulated on classical hardware using the PennyLane quantum machine learning platform. The proposed framework integrates Shapley Additive Explanations (SHAP) to improve model transparency and facilitate geological interpretation of the learned relationships. A total of 3213 depth-matched samples acquired from three wells in the Mpyo Field, Albertine Graben, Uganda, were used to develop and evaluate the model using five conventional well-log measurements: thermal neutron porosity (TNPH), effective porosity (PHIE), formation density (RHOZ), spectral gamma ray (SGR), and shale volume (VSH). Hyper-parameters of the QNN were optimized through systematic grid search and compared with a classical Group Method of Data Handling (GMDH), Random Forest (RF) and Support Vector Regression (SVR) models under identical experimental conditions. The simulated QNN achieved superior predictive accuracy, yielding training RMSE and MAE values of 0.019 and 0.0105, respectively, while demonstrating improved generalization on a blind test well compared with the benchmark model. SHAP analysis identified effective porosity as the dominant predictor of permeability, followed by thermal neutron porosity and spectral gamma ray, with feature contri-butions remaining consistent with established petrophysical principles. Bootstrap-based uncertainty analysis further demonstrated reliable prediction intervals across different reservoir intervals. Although implemented on a classical simulator rather than physical quantum hardware, the proposed framework illustrates the potential of quantum-inspired learning architectures for nonlinear permeability prediction while maintaining model inter-pretability for practical reservoir characterization
PROSS-assisted engineering enhances the thermal and storage stability of an AFB1-degrading superoxide dismutase
(Elsevier, 2026-09-26) Zhang, Yan; Zhang, Yiqian; Mwabulili, Fred
Aflatoxin B1, Superoxide dismutase, PROSS, Thermostability, Storage stability Aflatoxin B1 (AFB1) contamination threatens feed safety, highlighting the need for enzymatic detoxification strategies. In this study, an AFB1-degrading strain, HNGD-122, was isolated and identified as a Priestia mega-terium-related strain. Genome-guided mining identified a superoxide dismutase (SOD) with AFB1-degrading activity. To improve enzyme stability, PROSS-assisted design was applied, yielding the M163K mutant. M163K retained AFB1-degrading activity and exhibited enhanced thermostability, with half-lives at 70 and 80 ◦C extended to 32.2 and 20.7 h, respectively. It exhibited higher AFB1 degradation rates than wild-type (WT)-SOD during storage at 4 ◦C and following freeze-drying. Molecular dynamics simulations suggested that M163K maintained SOD–AFB1 interactions while improving thermostability through enhanced local stabilizing in-teractions. In zebrafish assays, it reduced developmental toxicity and oxidative stress after M163K treatment. In naturally contaminated corn flour and corn bran, M163K degraded approximately 66.5% and 61.9% of AFB1, respectively, supporting its potential for AFB1 control in corn-based feed ingredients.
An Interpretable MACD–SCDM Framework for Sepsis Risk Modeling from Longitudinal Laboratory Trajectories
(Electronics, 2026-06-26) Nyatega, Charles o; Adamu, Mohmmed
Early identification of sepsis from electronic health records (EHRs) remains challeng-
ing because longitudinal laboratory measurements are irregularly sampled, incomplete,
and dependent on the observation period. This study presents an interpretable tempo-
ral feature-engineering framework integrating Moving Average Convergence Divergence
(MACD)-based momentum descriptors with a Signal Correlation Decay Measure (SCDM)
representing short-term temporal persistence in longitudinal laboratory trajectories. Using
the MIMIC-IV Clinical Database Demo (v2.2), 275 hospital admissions from 100 unique
patients were screened. Temporal descriptors were derived from lactate, creatinine, white
blood cell count, platelet count, and international normalized ratio. Sepsis status was
defined at the hospital-admission level using prespecified ICD-9-CM and ICD-10-CM diag-
nosis codes, independently of laboratory predictors. Because reliable diagnosis timestamps
were unavailable, the task was formulated as admission-level sepsis risk classification
rather than prediction of time-stamped sepsis onset. Features were constructed over cu-
mulative 0–12 h, 0–18 h, and 0–24 h observation horizons. Lactate was used exclusively
as a predictor. HistGradientBoosting and Logistic Regression were evaluated using re-
peated patient-grouped cross-validation. Predictive performance varied across observation
horizons. Logistic Regression achieved the highest mean AUROC at 24 h (0.711 ± 0.125),
while HistGradientBoosting achieved 0.682 ± 0.139. Feature-set ablation indicated that
MACD-derived descriptors provided the most consistent incremental contribution relative
to conventional features, although their benefit varied by classifier and observation horizon.
Permutation importance further demonstrated horizon-dependent feature relevance, with
lactate-, INR-, platelet-, and WBC-derived descriptors among the highest-ranked predic-tors
TRAJECTORY TRACKING OF QUADROTOR UNMANNED AERIAL VEHICLE USING PARTICLE SWARM OPTIMIZATION BASED ADAPTIVE SUPERTWISTING SLIDING MODE CONTROL. .
(2026-05-20)
Quadrotor Unmanned Aerial Vehicles are increasingly used in search and res-cue, agriculture, mapping, service delivery, military surveillance, and power line inspection. However, their underactuated dynamics, coupled with parametric un certainties,unmodelled dynamics, and disturbances pose substantial challengesfor robust trajectory tracking and attitude stabilization.Among othersclassical controllers including Proportional Integral Derivatives, Model Predictive Control and Linear Quadratic Regulator offer acceptable perfor-mance but lack robustness under disturbances, and parameter variations. Sliding Mode Control overcomes key limitations of linear controllers with strong robust-ness, however, its main weakness is chattering caused by discontinuous switchingcontrol, which may excite unmodeled dynamics and wear actuator.This work develops a quaternion based dynamic model for quadrotor motion that captures both translational and rotational behavior. Unlike many existing models, the proposed formulation includes unmodeled effects such as aerodynamicdrag and propeller induced forces, resulting in a more realistic and more accurate representation of quadcopter flight dynamics. An Adaptive Super Twisting Sliding Mode Controller (ASTSMC) based on quater- nion modeling was designed such that quaternion representation ensures smooth attitude tracking without gimbal lock, while Super Twisting reduces chattering and improves robustness.M Particle Swarm Optimization (PSO) was used to tune the sliding surface gains of the ASTSMC, reducing attitude objective values by 4.43%–27.22% with respectto initial global best value and improving position accuracy by up to 10.62%. The Best Global value drop from 10.49 to 7.66 reflects a 26.97% reduction in combined tracking error and control effort. Gains tuned by PSO yield faster convergence, smoother response, and better performance than manual tuning. The proposed PSO ASTSMC significantly enhances quadrotor trajectory track-ing by reducing attitude errors by up to 27% and improving position accuracy by 0.61%–10.62% relative to a conventional PID and backstepping sliding mode control under identical simulation conditions.