Sobol Methods Research Articles

Global sensitivity analysis of Open Systems Pharmacology Suite physiologically based pharmacokinetic models.

Sensitivity analyses are important components of physiologically based pharmacokinetic (PBPK) model development and are required by regulatory agencies for PBPK submissions. They assess the impact of parametric uncertainty and variability on model estimates, aid model optimization by identifying parameters requiring calibration, and enable the testing of assumptions within PBPK models. One-at-a-time (OAT) sensitivity analyses quantify the impact on a model output in response to changes in a single parameter while holding others fixed. Global sensitivity analysis (GSA) methods provide more comprehensive assessments by accounting for changes in all uncertain or variable parameters, though at a higher computational cost. This tutorial article presents a software package for conducting both OAT and GSA of PBPK models built in the Open Systems Pharmacology (OSP) Suite. The tool is accessible through either an R script or a graphical user interface, and the outputs consist of sensitivity metrics of pharmacokinetic (PK) parameters, such as Cmax and AUC, evaluated with respect to model input parameters. Results are formatted according to regulatory standards. The OAT analysis methods comprise two-way local sensitivity analyses and probabilistic uncertainty analyses, whereas the GSA methods include the Morris, Sobol, and EFAST methods. These analyses can be conducted on single PBPK models or pairs of models for the evaluation of the sensitivity of PK parameter ratiosin drug-drug interaction studies. The practical application of the package is demonstrated through three illustrative case studies.

Creep reliability assessment of structural components at elevated temperatures considering the time dependent feature of representative stress

Creep reliability assessment of structural components at elevated temperatures is essential to guarantee the long-term safe operation of the system. Current studies are limited to continuum damage mechanics methods at the material level, while the reliability assessment method for creep design at the component level is rarely reported. In this work, the framework for creep reliability assessment of structural components is extended, where the time dependent feature of the representative stress is included. The effect of the time dependent feature of the representative stress on creep reliability assessment is discussed. Sensitivity analyses of material parameters on creep reliability assessment results are conducted based on the Sobol and Morris global methods. Results indicate that for the same creep design life, the component presents a higher failure probability when the time dependent feature of the representative stress is considered. Parameters D and d in the creep rupture life equation have more significant effects on creep rupture life than other parameters for the case studied.

Comprehensive Sensitivity Analysis of Cost-Benefit Analysis Variables for Transport Interventions

Cost-benefit analysis (CBA) is the universally applied tool to assess economic viability in assisting decisions on transport investments. Its framework is heavily influenced by the numerous variables it considers through estimating and valuing the intervention’s effects. This paper – utilising the authors’ previously implemented CBA test environment – comprehensively analyses the sensitivity of significant variables of three typical CBA models for transport interventions (road, rail and urban) to understand the prevailing appraisal approach better and to help focus on further methodological improvements. Morris and Sobol methods were selected to study the global sensitivity of and the relations between the input parameters of the models. The sensitivity test of the three analysed models provided similar results regarding which variables are most influential in CBAs. Input variables such as the investment cost, the economic discount rate, forecasted GDP changes and specific elasticities to these GDP changes often have a firm but mostly linear effect. Value of time, vehicle operating cost and mode choice-related parameters such as car availability, car occupancy rate, level of service indicators (e.g. frequency of service) and potential to induce travel demand (proxied by a ‘no travel’ parameter) are inputs with considerable linear effects and greater interactive effects.

New sexually transmitted HIV infections from 2016 to 2050 in Guangdong Province, China: a study based on a dynamic compartmental model

BackgroundIn Guangdong Province, China, there is lack of information on the HIV epidemic among high-risk groups and the general population, particularly in relation to sexual transmission, which is a predominant route. The new HIV infections each year is also uncertain owing to HIV transmission from men who have sex with men (MSM) to women, as a substantial proportion of MSM also have female sexual partnerships to comply with social demands in China.MethodsA deterministic compartmental model was developed to predict new HIV infections in four risk groups, including heterosexual men and women and low- and high-risk MSM, in Guangdong Province from 2016 to 2050, considering HIV transmission from MSM to women. The new HIV infections and its 95% credible interval (CrI) were predicted. An adaptive sequential Monte Carlo method for approximate Bayesian computation (ABC-SMC) was used to estimate the unknown parameter, a mixing index. We calibrated our results based on new HIV diagnoses and proportions of late diagnoses. The Morris and Sobol methods were applied in the sensitivity analysis.ResultsNew HIV infections increased during and 2 years after the COVID-19 pandemic, then declined until 2050. New infections rose from 8,828 [95% credible interval (CrI): 6,435–10,451] in 2016 to 9,652 (95% CrI: 7,027–11,434) in 2019, peaking at 11,152 (95% CrI: 8,337–13,062) in 2024 before declining to 7,084 (95% CrI: 5,165–8,385) in 2035 and 4,849 (95% CrI: 3,524–5,747) in 2050. Women accounted for approximately 25.0% of new HIV infections, MSM accounted for 40.0% (approximately 55.0% of men), and high-risk MSM accounted for approximately 25.0% of the total. The ABC-SMC mixing index was 0.504 (95% CrI: 0.239–0.894).ConclusionsGiven that new HIV infections and the proportion of women were relatively high in our calibrated model, to some extent, the HIV epidemic in Guangdong Province remains serious, and services for HIV prevention and control are urgently needed to return to the levels before the COVID-19 epidemic, especially in promoting condom-based safe sex and increasing awareness of HIV prevention to general population.

Quantitative risk analysis and parameter sensitivity evaluation of wellbore instability in poroelastic media considering uncertainty of geomechanical parameters

The randomness and fuzziness of geomechanical parameters make it difficult to directly determine the wellbore collapse and fracture pressures, and thus show significant influence on the risk assessment of drilling operations. Therefore, to effectively analyze the risk of wellbore instability in deep formation, this paper develops the evaluation models of wellbore stability for the vertical well drilled through the poroelastic rock formation subjected to borehole stresses under different time domains, including the instantaneous, modified instantaneous, short-time, long-time, and elastic ones. Meanwhile, based on the Monte Carlo method, one compares the difference between Taylor expansion and the Rosenbluth method in quantifying the uncertainty of the evaluation models. One analyzes the risk of wellbore instability under the condition that the geomechanical parameters satisfy the uniform, triangular, and normal distribution characteristics. In addition, different from the single factor sensitivity evaluation in previous uncertainty analysis, one introduces Morris and Sobol methods to conduct the comprehensive analysis of local and global sensitivity of geomechanical parameters. The main results show that the collapse pressure determined by the different cases of borehole stresses presents in descending order of short-time, long-time, and instantaneous. However, the fracture pressure shows in ascending order of long-time, instantaneous, and short-time cases. Besides, the different distribution characteristics of geomechanical parameters lead to the relationship of safe mud weight window in descending order of normal distribution, triangular, uniform distributions, and the uncertainty characteristics caused by the normal distribution of geomechanical parameters can be effectively quantified by Taylor expansion and Rosenbluth method. The strong sensitivity parameters affecting wellbore collapse and fracture are the maximum and the minimum horizontal in situ stresses, respectively. These two parameters present a strong difference in global sensitivity and local sensitivity analysis, which is manifested by the comprehensive effect of global parameters. The influence of the maximum horizontal in-situ stress on borehole collapse is weakened to a certain extent, while wellbore tensile fracturing is strengthened. On the contrary, the minimum horizontal in-situ stress is characterized by enhancing the borehole collapse whereas weakening the tensile fracturing. Besides, the influence degree increases with the increasing of parameter uncertainty.

Toward presenting an ensemble meta‐model for evaluation of pozzolanic mixtures incorporating industrial by‐products

AbstractUnsustainable development in the industry has increased the challenge of proper disposal of industrial by‐products and has caused environmental pollution on a large scale. Furthermore, minimizing the carbon dioxide emissions due to industrial and environmental activities is a crucial challenge to controlling the global warming crisis and climate change. The use of industrial by‐product‐based supplementary cementitious materials (SCMs) can reduce fossil fuel consumption through the reduction in cement production and contribute to suitable waste management. In this research, a meta‐model was developed using integration of the decision tree‐based method named reduced error pruning tree (REPtree) and ensemble approaches (boosting and bagging) to improve the pruning phase for strength simulation of concrete mixtures incorporating industrial by‐products. Then, ensemble multiple learning algorithms were applied to develop a meta‐model using the artificial neural network and Gaussian process regression. A comprehensive experimental‐based database incorporating 3836 records from short‐term to long‐term compressive strength (CS) was collected from the literature. Performance measures present that the proposed ensemble boosting‐REPtree model outperforms other benchmark and ensemble models with a correlation coefficient of 0.960 and root mean square error of 7.884 MPa. Also, uncertainty evaluation was done based on the Monte Carlo simulation algorithm in the modeling process. The FAST and SOBOL methods, as global sensitivity analysis, presented water‐to‐binder (W/B) ratio, curing time, and SCMs to binder ratio (SCM/B) as critical components to estimate the CS of concrete mixtures. The computational simulation of the influential parameters on concrete mixtures incorporating industrial by‐products revealed that the experimental factors yielding the highest CS of the mixtures were calculated as follows: cement content of 584 kg/m3, W/B ratio of 0.17, superplasticizer‐to‐cement ratio of 1.88%, and SCM/B ratio of 0.15.

Global sensitivity analysis of sound attenuation in double-wall system with porous layer

This paper investigates the acoustic performance of a double-wall system with a porous layer and presents a global sensitivity analysis of sound attenuation. The transfer matrix method is first applied to predict sound transmission through the structure. This method offers a relatively simple and cost-effective way to model the complex acoustic interactions in these systems and provides rich high-frequency information. The method represents sound wave propagation for each layer using a transfer matrix that depends on the thickness and physical characteristics of each material, and interface matrices are introduced to consider boundary conditions between adjacent layers. The poroelastic layer is modeled using the Biot-Allard approach with nine parameters. Then a global sensitivity analysis is performed using Morris and Sobol methods to identify the parameters that have a significant impact on sound transmission loss. The Morris method is used first to eliminate the parameters that have the least impact, and the Sobol method is then employed to analyze the remaining parameters and their interactions in more detail. The study focuses on eleven parameters, including all the physical parameters of the foam, and thicknesses of plates, cavity, and foam. The results of the sensitivity analysis indicate that geometrical parameters such as the thickness of different layers have the most significant impact on the sound transmission loss response in the lower frequency range. In contrast, foam properties such as flow resistivity have more influence in the higher frequency range. Overall, this study is the first to apply sensitivity analysis methods to the problem of sound transmission through double wall systems with porous layers, providing valuable insights into the system’s behavior and its design optimization.

Application of sensitivity analysis in extension, inflation, and torsion of residually stressed circular cylindrical tubes

This paper deals with applying two main sensitivity analysis (SA) methods, namely, the Sobol method and the Fourier Amplitude Sensitivity Test (FAST) method on the problem of mixed extension, inflation, and torsion of a circular cylindrical tube in the presence of residual stress. The mechanical side of the problem was previously proposed by Merodio & Ogden (2016). The input parameters in the form of the initial cylinder geometry, the amount of the residual stress, the azimuthal stretch, the axial elongation, and the torsional strain are distributed according to three probability distribution methods, namely the uniform, the gamma, and the normal distribution. In the present work, through applying Sobol and FAST methods, the most influential factors among input parameters on the output variable have been determined. The assessment of our results is then determined by the computation of bias and standard deviation of Sobol and FAST indices for each input parameter in the model.

Feasibility of neural network metamodels for emulation and sensitivity analysis of radionuclide transport models

In this paper we compare the outputs of neural network metamodels with numerical solutions of differential equation models in modeling cesium-137 transportation in sand. Convolutional neural networks (CNNs) were trained with differential equation simulation results. Training sets of various sizes (from 5120 to 163,840) were used. First order and total order Sobol methods were applied to both models in order to test the feasibility of neural network metamodels for sensitivity analysis of a radionuclide transport model. Convolutional neural networks were found to be capable of emulating the differential equation models with high accuracy when the training set size was 40,960 or higher. Neural network metamodels also gave similar results compared with the numerical solutions of the partial differential equation model in sensitivity analysis.

Sensitivity analysis methodology to identify the critical material properties that affect the open hole strength of composites

Aeronautical industries are concerned about generating efficient and cost-effective statistical design allowables of composite laminate responses. Within this framework of uncertainty quantification and management (UQ & M), accounting for the various uncertainties is crucial to calculate the design allowables which, in turn, adds confidence to the design and certification process. In this context, as a precursor to UQ & M, we establish a global sensitivity analysis framework to identify the material properties that influence the laminate responses. In this study, an open hole tensile (OHT) test configuration is considered. In a first step, sensitivity analysis based on an analytical model is performed using Sobol, FAST, and Morris methods, whereas in a second step, sensitivity analysis based on a detailed 3D finite element model (FEM) incorporating damage is performed using Morris. The paper compares different sensitivity approaches in terms of computational cost and results. Further, different case studies (varied notched hole sizes, width-to-diameter ratios, laminate stacking sequences, etc.) are performed to study how the sensitive parameters vary for different configurations. The study identified the Morris method, along with the selected hyper parameters, as being suitable for FEM based sensitivity analysis due to its low computational cost. Five material properties (out of 23) were identified as influencing the open hole laminate tensile strength. Using the FE simulations performed for the Morris analysis, a first estimation of the design allowables (A and B basis values) on the OHT strength is also presented. The results from this paper contribute to the next step of UQ&M, where only the identified sensitive material properties can be considered for detailed experimental characterization or for building meta models to calculate the design allowables. A framework like this, complements experimental campaign to reduce the associated costs and knock-down factors, thus shortening the certification process.

Response Surface Methodology for Copper Flotation Optimization in Saline Systems

Response surface methodology (RSM) is one of the most effective tools for optimizing processes, and it has been used in conjunction with the Analysis of Variance (ANOVA) test to establish the effect of input factors on output factors. However, when this methodology is used in mineral flotation, its polynomial model usually performs poorly. An alternative is to use artificial neural networks (ANNs) in such situations. Within this context, the ANOVA test is not the best option for these model types; moreover, it requires statistical assumptions that are difficult to satisfy in flotation. This work proposes replacing the polynomial model of the RSM with ANNs and the Sobol methods to determine the influential input factors instead of the ANOVA test. This proposal is applied to two porphyry copper ores with a high content of pyrite, clay, and dilution media. In addition, this study shows how other computational intelligence techniques, such as swarm intelligence, can be incorporated into this type of problem to improve the learning process of ANNs. The results gave an adjustment of over 0.98 for R2 using ANNs, in comparison to values of around 0.5 when the polynomial model of RSM was utilized. On the other hand, the application of Global Sensitivity Analysis (GSA) identified the aeration rate and P80 size as the most influential variables in copper recovery under the conditions studied. Additionally, we identified significant interactions that affect the recovery of copper, with the interactions between the aeration rate, frother concentration, and P80 size being the most important.

The effect of different sampling method integrated in NSGA II optimization on performance and emission of diesel/hydrogen dual-fuel CI engine

Recent advances in computational tools and techniques have enabled engineers to tackle single-step and iterative complex computer-aided engineering (CAE) problems. They have also helped develop CAE optimization techniques capable of driving design parameters towards regions where selected system characteristics can be further improved. The present work was conducted in two steps; In the first step, the effects of adding hydrogen to diesel fuel in a turbocharged direct injection (TDI) compression ignition (CI) Audi engine was analyzed through 1D simulation in AVL BOOST. With no significant modifications to the engine, the analysis was carried out at 1900 rpm and 2500 rpm under 0.8 MPa nominal loading conditions. Hydrogen was injected with air at 30 L/min through the intake manifold before the turbocharger. The present work considered all the engine components that are thermodynamically involved in the combustion process. The second analysis step coupled the simulated model with modeFRONTIER to study the optimum parameter values for engine performance and emission and determined the effects of various input variables on these parameters. Optimization was carried out by adopting hybrid algorithms to design experiments and the non-dominated sorting genetic algorithm II (NSGA_II). The response level predictive capability of the Taguchi and SOBOL methods increases the convergence speed of the NSGA_II. The results from the first analysis step indicated that the hydrogen diesel blend reduced brake specific fuel consumption (BSFC), engine power and torque, CO2 and NOx emissions, and increased entropy. The results from the second step suggested that in multivariable optimization, the peak pressure and NOx emissions were affected mainly, by combustion initiation time and the hydrogen added to the diesel fuel, respectively. Finally, it was found that hydrogen injection at 20 mg/min resulted in the best engine performance and emission.

Probabilistic failure prediction in a double composite cantilever beam with single and double source uncertainty

Predicting failure in composite materials under service loading conditions has been challenging due to the non-uniform mechanical properties arising from the composite fabrication process. Including these uncertainties in the analysis becomes critical. The probabilistic approach plays a vital role in making the design less conservative and anticipates the risk associated with the design incorporating the uncertainties. In this work, metamodels such as support vector machines, radial basis function, and logistic regression in conjunction with Latin hypercube, Sobol, and Halton sequence sampling methods were used to calculate the failure probability in the carbon fibre/epoxy-based composite material. Here, the composite plates were fabricated using the vacuum-assisted resin transfer molding (VARTM) process. The variation in the fibre-volume fraction was evaluated at different sites of the composite plate. Then, the effective orthotropic properties of the composite for various fibre-volume fractions have been numerically computed by the homogenisation method using periodic boundary conditions. A double cantilever composite beam problem was considered to predict the failure probability by including the uncertainties in single-source — fibre-volume fraction and double-source — fibre-volume fraction and fracture toughness. At the end, a study to ascertain the metamodels stability was presented to demonstrate the accuracy and effectiveness of the proposed approach.

Identifying influential parameters of a multi-species fish size spectrum model for a northern temperate lake through sensitivity analyses

Ecosystem-based approaches that take species interactions into account have shifted to the forefront of fisheries modelling and management in recent years. As a result, multispecies size spectrum models have been increasingly used to explore impacts of fishing on marine community dynamics. The use of these models, which has been facilitated by the development of the R package mizer, requires the estimation of species-specific parameters related to growth, reproduction, and feeding. These parameters, which may be estimated from imperfect information, may contribute to model uncertainty and thus reduce the value of information available for management purposes. In this study of a freshwater fishery, we conduct a comprehensive global sensitivity analysis pairing the Morris and Sobol methods to identify life-history parameters having the largest influence on model outputs. Here, we focus on (i) the size spectrum slope, (ii) the scatter around the linear relationship of the size spectrum, (iii) total biomass, and (iv) species diversity. We found that parameters relating to growth, namely the von Bertalanffy growth coefficient and asymptotic mass, had the greatest influence on our size spectrum model results. This was particularly true for top predators and the most abundant species. Our results suggest that estimation of growth parameters of top predators be given priority to reduce uncertainty in model output, and ultimately, fisheries management.

Flattening Layer Pruning in Convolutional Neural Networks

The rapid growth of performance in the field of neural networks has also increased their sizes. Pruning methods are getting more and more attention in order to overcome the problem of non-impactful parameters and overgrowth of neurons. In this article, the application of Global Sensitivity Analysis (GSA) methods demonstrates the impact of input variables on the model’s output variables. GSA gives the ability to mark out the least meaningful arguments and build reduction algorithms on these. Using several popular datasets, the study shows how different levels of pruning correlate to network accuracy and how levels of reduction negligibly impact accuracy. In doing so, pre- and post-reduction sizes of neural networks are compared. This paper shows how Sobol and FAST methods with common norms can largely decrease the size of a network, while keeping accuracy relatively high. On the basis of the obtained results, it is possible to create a thesis about the asymmetry between the elements removed from the network topology and the quality of the neural network.

Reliability and Reliability-based Sensitivity Analyses of Steel Moment-Resisting Frame Structure subjected to Extreme Actions

The ground external columns of buildings are vulnerable to the extreme actions such as a vehicle collision. This event is a common scenario of buildings' damages. In this study, a nonlinear model of 2-story steel moment-resisting frame (SMRF) is made in OpenSees software. This paper aims investigating the reliability analysis of aforementioned structure under heavy vehicle impact loadings by Monte Carlo Simulation (MCS) in MATLAB software. To reduce computational costs, meta-model techniques such as Kriging, Polynomial Response Surface Methodology (PRSM) and Artificial Neural Network (ANN) are applied and their efficiency is assessed. At first, the random variables are defined. Then, the sensitivity analyses are performed using MCS and Sobol's methods. Finally, the failure probabilities and reliability indices of studied frame are presented under impact loadings with various collision velocities at different performance levels and thus, the behavior of selected SMRF is compared by using fragility curves. The results showed that the random variables such as mass and velocity of vehicle and yield strength of used materials were the most effective parameters in the failure probability computation. Among the meta-models, Kriging can estimate the failure probability with the least error, sample number with minimum computer processing time, in comparison with MCS.

Global sensitivity analysis of VISSIM parameters for project-level traffic emissions: a case study at a signalized intersection

ABSTRACT Combining traffic micro-simulation models and emission models is a primary method to estimate traffic emissions. However, there has been limited research on the impact of traffic micro-simulation model parameters on simulator outputs, which are critical to emissions calculations. Based on combining VISSIM and MOVES, this study uses the Morris and Sobol methods to explore the impact of parameters in VISSIM on operating mode distribution and travel time distribution. Taking an urban signal control intersection with three traffic scenarios in Chengdu, China as an example, this study verifies the methods’ feasibility. Apart from these parameters, which have been proved to be necessary calibration parameters, including the desired speed distribution, the desired acceleration function, and the desired deceleration function, an additional 24 parameters related to simulation setting and driving behaviour models are selected as the initial parameters. The number of interaction objects, maximum look-ahead distance, average standstill distance, additive part of safety distance, and safety distance reduction factor close to a stop line, are considered to be the important parameters for this case study. The impact of these five parameters on the bins of operating mode distribution and travel time distribution are further analyzed with One-at-a-time, and these parameters are compared with those reported in previous studies. It is concluded that the important parameters selected in this study are reasonable and can support the calibration of VISSIM parameters for this case’s traffic emissions.

Evaluation of capabilities of different global sensitivity analysis techniques for building energy simulation: experiment on design variables

Abstract The objective of this study is to investigate the capabilities of different global sensitivity analysis methods applied to building performance simulation, i.e. Morris, Monte Carlo, Design of Experiments, and Sobol methods. A single-zone commercial building located in Florianópolis, southern Brazil, was used as a case study. Fifteen inputs related to design variables were considered, such as thermal properties of the construction envelope, solar orientation, and fenestration characteristics. The performance measures were the annual heating and cooling loads. It was found that each method can provide different visual capabilities and measures of interpretation, but, in general, there was little difference in showing the most influent and least influent variables. For the heating loads, the thermal transmittances were the most influent variables, while for the cooling loads, the solar absorptances stood out. The Morris method showed to be the most feasible method due to its simplicity and low computational cost. However, as the building simulation model is still complex and non-linear, the variance-based method such as the Sobol is still necessary for general purposes.

CO2-prepad injection EOR simulation and sensitivity analysis considering miscibility and geomechanics in tight oil reservoirs

Injecting CO2 pre-pad into tight reservoirs before hydraulic fracturing is a promising way to improve the productivity of single well. In this method, CO2 should be injected into the formation at a high rate. Also the miscibility of CO2 and oil with the changes of intricate geomechanics during this process are complex. Meanwhile, the geological and operating parameters have a dramatic influence to oil production and is lack of systematic research. In this study, a fluid model of crude oil was established to evaluate the impact of CO2 on oil volume and viscosity. Then, a compositional numerical model coupled miscibility and geomechanics were set up to analyze enhanced oil recovery (EOR) potential under different miscible degree with in-situ stress field. Ultimately, in accordance with multiple nonlinear regression production forecast models, Sobol, FAST (Fourier Amplitude Sensitivity Test), and DGSM (derivation-based Global Sensitivity Measure) methods were utilized to study the main factors affecting cumulative production. The results shown that CO2 almost has the same effect on the volume of these tight oil and has the most significant effect on the viscosity of crude oil we studied by comparing. The production of miscible flooding is 38.2% and 16.4% higher than that of immiscible and near-miscible flooding, respectively. The formation were compacted during production especially in the early stage, hence the production of the model with geomechanics was 23.3% less than the model of non-geomechanics. The first-order sensitivity indexes of bottom-hole flow pressure, reservoir thickness and fracture conductivity are about 0.808, 0.176 and 0.004 respectively obtained by Sobol method, consistent with two other methods. This result shows that bottom-hole flow pressure, reservoir thickness and fracture conductivity are the most important factors affecting production, and bottom-hole flow pressure has the most significant impact on production. This work investigates how miscibility and geomechanics effect on production as well as clarified the factors affecting production for CO2 pre-pad injection, providing vital insights into operating in tight reservoirs for oil field.

Considerations and Caveats when Applying Global Sensitivity Analysis Methods to Physiologically Based Pharmacokinetic Models

Three global sensitivity analysis (GSA) methods (Morris, Sobol and extended Sobol) are applied to a minimal physiologically based PK (mPBPK) model using three model drugs given orally, namely quinidine, alprazolam, and midazolam. We investigated how correlations among input parameters affect the determination of the key parameters influencing pharmacokinetic (PK) properties of general interest, i.e., the maximal plasma concentration (Cmax) time at which Cmax is reached (Tmax), and area under plasma concentration (AUC). The influential parameters determined by the Morris and Sobol methods (suitable for independent model parameters) were compared to those determined by the extended Sobol method (which considers model parameter correlations). For the three drugs investigated, the Morris method was as informative as the Sobol method. The extended Sobol method identified different sets of influential parameters to Morris and Sobol. These methods overestimated the influence of volume of distribution at steady state (Vss) on AUC24h for quinidine and alprazolam. They also underestimated the effect of volume of liver (Vliver) for all three drugs, the impact of enzyme intrinsic clearance of CYP2C9 and CYP2E1 for quinidine, and that of UGT1A4 abundance for midazolam. Our investigation showed that the interpretation of GSA results is not straightforward. Dismissing existing model parameter correlations, GSA methods such as Morris and Sobol can lead to biased determination of the key parameters for the selected outputs of interest. Decisions regarding parameters’ influence (or otherwise) should be made in light of available knowledge including the model assumptions, GSA method limitations, and inter-correlations between model parameters, particularly in complex models.Graphical abstract