Lixoft offers tutorials, webinars on real test cases and “feature of the week” videos.
- Published tutorial
- Population modeling with Monolix
- PK models
- Specific data type
- Simulations with Simulx
- Population modeling with Monolix
- Features of the week
- Data filtering
- Plots in MonolixSuite
- Software preferences
- Input and output files
- Building the structural model
- Building the statistical model
- Random effects for individual parameters
- Covariate effects for individual parameters
- Residual error
- Selection of statistical model
- Tasks of Monolix and modeling workflow
- Algorithms and results
- Estimating population parameters
- Estimating individual parameters
- Diagnostic plots
- Examples of specific models or data (continuous data only)
- Modeling non-continuous data
- Traynard, P, Ayral, G, Twarogowska, M, Chauvin, J (2020). Efficient Pharmacokinetic Modeling Workflow With the MonolixSuite: A Case Study of Remifentanil. CPT Pharmacometrics Syst Pharmacol. https://doi.org/10.1002/psp4.12500.This tutorial presents a step‐by‐step pharmacokinetic (PK) modeling workflow using MonolixSuite 2019, including how to visualize the data, set up a population PK model, diagnose and improve the model incrementally, perform a covariate search, and keep track of the different runs in the workflow.
- Ayral Geraldine, Si Abdallah Jean-François, Magnard Claude, Chauvin Jonathan (2021). A novel method based on unbiased correlations tests for covariate selection in nonlinear mixed effects models: The COSSAC approach CPT Pharmacometrics Syst Pharmacol. https://doi.org/10.1002/psp4.12612.This tutorial presents a novel stepwise method to build a covariate model based on statistical tests between individual parameters sampled from their conditional distribution and the covariates. This strategy, called the COnditional Sampling use for Stepwise Approach based on Correlation tests (COSSAC), makes use of the information contained in the current model to choose which parameter‐covariate relationship to try next. In this article, we detail the COSSAC method and its implementation in Monolix, and evaluate its performance.
Population modeling with Monolix
Warfarin case study : This video case study shows a simple PK modeling workflow in Monolix2018, with the example of warfarin. It explains the main features and algorithms of Monolix, that guide the iterative process of model building: from validating the structural model to adjusting the statistical model step-by-step. It includes picking a model from the libraries, choosing initial estimates with the help of population predictions, estimating parameters and uncertainty, and diagnosing the model with interactive plots and statistical tests.
- Tobramycin case study : This case study presents the modeling of the tobramycin pharmacokinetics, and the determination of a priori dosing regimens in patients with various degrees of renal function impairment. It takes advantage of the integrated use of Datxplore for data visualization, Mlxplore for model exploration, Monolix for parameter estimation and Simulx for simulations and best dosing regimen determination.
- Remifentanil case study : Remifentanil is an opioid analgesic drug with a rapid onset and rapid recovery time. It is used for sedation as well as combined with other medications for use in general anesthesia. It is given in adults via continuous IV infusion, with doses that may be adjusted to age and weight of patients. This case-study shows how to use Monolix2018 to build a population pharmacokinetic model for remifentanil in order to determine the influence of subject covariates on the individual parameters.
An updated version of this tutorial with Monolix2019 has been published in CPT Pharmacometrics Syst Pharmacol: https://doi.org/10.1002/psp4.12500
- Veralipride case study : Multiple peaking in plasma concentration-time curves is not uncommon, and can create difficulties in the determination of pharmacokinetic parameters. For example, double peaks have been observed in plasma concentrations of veralipride after oral absorption. While multiple peaking can be explained by different physiological processes, in this case site-specific absorption has been suggested to be the major mechanism. In this webinar we explore this hypothesis by setting up a population PK modeling workflow with the MonolixSuite 2018.
- PK model development and covariate exploration (Maryland 1) : The aim of this tutorial is to develop a population PK model and explore covariate relationships of a hypothetical test drug used in the treatment of atrial fibrillation.
- PK model development for a multi-dose study (Maryland 2) : The aim of this tutorial is to develop a population PK model for a multiple dosing trial study of a hypothetical test drug used in the treatment of diabetes.
- PK/PD modeling using the simultaneous, sequential or intermediate approach (Maryland 3) : The aim of this tutorial is to show how to develop a pharmacokinetic-pharmacodynamic (PKPD) model. It is based on a clinical study of a hypothetical test drug used in the treatment of immunosuppression and as an anti-inflammatory agent. Three different approaches are described in detail: sequential, simultaneous (joint) and intermediate (population sequential).
- Inter-occasion variability and effect of guar gum on alcohol concentration in blood : In this case study, the MonolixSuite is used to analyze and model the PK of alcohol, measured in two different occasions where subjects have taken or not a dietary additive of guar gum. The case study focuses in particular on the modeling of inter-occasion variability and the effect of guar gum on the bioavailability. Several applications of the MonolixSuite are used: Datxplore for data exploration, PKanalix for NCA and CA, Monolix for population modeling, and Sycomore to compare several Monolix runs.
Specific data types
- Concentration-QTc modeling with MonolixSuite : Model based studies of concentration-QT data is a primary analysis in the proarrhythmic risk assessment (ICH E14 Guidance). Performed as a standard part of modelling and simulation at different stages of drug development, c-QTc analysis in MonolixSuite provides early, detailed and reliable insight necessary for companies to better manage their pipelines. In this webinar you will learn what are the current c-QTc modelling objectives and approaches and how to perform FDA recommended c-QTc analysis in MonolixSuite.
- QSP modeling with MonolixSuite – Session 1 :This video gives an introduction to quantitative systems pharmacology (QSP) and shows a step-by-step case study for an FAAH inhibitor highlighting the typical workflow and good practices. Model implementation in Mlxtran language, parameter estimation in Monolix, and model simulation to perform predictions in Simulx are covered. Sensitivity analysis and profile likelihood methods are also presented and illustrated.
- QSP modeling with MonolixSuite – Session 2 : This second session focuses on the development of a model for a cholesterol-lowering anti-PCSK9 drug. Three different models of varying complexity (typical PK/PD, mechanistic PK/PD and QSP) are developed and compared in terms of prediction capabilities and ease of use.
- Tumor growth modeling with MonolixSuite – Session 1 : This video presents common tumor growth and tumor growth inhibition models from the literature, with guidelines to use them in Monolix. It shows a step-by-step case study on combination therapy in lung cancer xenografts.
- Tumor growth modeling with MonolixSuite – Session 2 : This second session explores tumor growth models that consider the emergence of resistance or other additional features such as angiogenesis or immune dynamics. It shows a modeling case study on prostate cancer tumor progression and survival.
- New Library of Tumor Growth Inhibition Models in Monolix : This one-hour webinar from March 2021 is a more condensed version of the information given in the two webinars above, presenting the new modular library of tumor growth inhibition models implemented in MonolixSuite 2020R1. This webinar includes an introduction of the different models available in the library, detailed guidelines to choose your model based on the shape of your data, and short real-world examples.
- Longitudinal Model-Based Meta-Analysis (MBMA) with Monolix Suite: Longitudinal model-based meta-analysis (MBMA) models can be implemented using the MonolixSuite. These models use study-level aggregate data from the literature and can usually be formulated as non-linear mixed-effects models in which the inter-arm variability and residual error are weighted by the number of individuals per arm. We exemplify the model development and analysis workflow of MBMA models in Monolix using a real data set for rheumatoid arthritis, following publication from Demin et al (2012). In the case study, the efficacy of a drug in development (Canakinumab) is compared to the efficacy of two drugs already on the market (Adalimumab and Abatacept). Simulations using Simulx were used for decision support to see if the new drug has a chance to be a better drug.
- Analysis of time-to-event data : Within the MonolixSuite, the mlxtran language allows to describe and model time-to-event data using a parametric approach. This page provides an introduction on time-to-event data, the different ways to model this kind of data, and typical parametric models. A library of common TTE models is also provided. Two modeling and simulation workflows illustrate this approach, using two TTE data sets.
Simulations with Simulx
- Optimizing Sample Size of a Phase III Trial with Simulx : What if we could reduce the size of a phase III study from 400 to 80 subjects and make it twice as short? Thanks to model-informed clinical trial design, it’s possible! Watch an inspiring story based on real-world case study data and learn about Simulx at the same time. In an hour, we will take you from the modeling of phase II studies in Monolix to the estimation of a phase III study power in Simulx to finally optimizing sample size and study duration. Example scripts to script Simulx from R as done in this webinar can be found here.
- From Monolix to Simulx: how to explore new scenarios : You have built a model and now you wonder how to compare new dosing regimens for the next trial, calculate the expected power of a study and select the most successful strategy? Re-wire your thinking to effortless exploration of new scenarios with Simulx GUI – intuitive, flexible, and powerful application to simulate and compare countless strategies – and switch your focus to analysis and decision-making. Watch this video to learn using Simulx with Monolix projects and discover features that bring the best insight from simulations. It shows step-by-step how to simulate different groups, define target outcomes and assess the uncertainty.
- Setting up a simulation from scratch with Simulx : No matter how you have developed your population PK/PD or QSP model, Simulx-GUI is an intuitive, flexible and powerful application to simulate new situations and answer “what if” questions. In this webinar, we explain how to setup a simulation from scratch by defining the model, parameters, treatments and outputs. A special focus will be given on the model writing in mlxtran language, with examples of translations from Nonmem and literature models.
- What’s new in MonolixSuite2020: PK case study on two formulations : This webinar shows the new features of MonolixSuite2020 with a PK case study example in which two theophylline formulations are compared with NCA and population modeling, and new dosing regimens are simulated.
- What’s new in MonolixSuite2021: Parent-Metabolite case study : This webinar shows the new features of MonolixSuite2021 with a parent-metabolite case study. In 2021, Monolix is enhanced with the long-awaited automated initialization of parameters for any model, a new parent-metabolite library and more analytical solutions to speed up calculations. In Simulx, accounting for uncertainty in the predictions is now a piece of cake. Our refreshed MlxEditor now fully-integrated in other apps will make it easier for you to refine your own models. Last but not least, PKanalix continues to evolve with a brand new bioequivalence module.
Features of the week
- #105: Using data set filters (part 1) : From version 2020R1 onward, it is possible to apply filters on your data set to work with a subset of it only. This video shows you how to proceed.
- #106: Using data set filters (part 2) : In this video we explain how to combine several filter actions using unions and intersections.
Plots in MonolixSuite
- #01 The beautiful interactive plots of Monolix2018
- #03 Stratifying plots in Monolix : Stratifying plots: a simple but impressively powerful new feature of Monolix for data exploration and model diagnosis.
- #11 Customizing the plots : Are you looking for publication-ready diagnostic plots? Monolix allows you to change the appearance of the plots to save high-quality figures.
- #24 Modifying the plot layout : Find out how to fully customize your plot layout in Monolix.
- #60 Id highlighting across plots :Monolix now includes an extended feature to highlight ids across plots! Watch this video to see what a great help it is to diagnose quickly your model.
- #78 Creating a custom theme for Monolix plots : You can create one or several custom themes and apply it to all Monolix plots when appropriate, and even share them with your colleagues. It is thus easy to generate ready-to-print figures that meet your personal requirements.
- #100:Trend lines : Can you imagine model building without data visualization? Neither can we, and this is why we keep improving it. Watch this video to see how the new trendlines for observed data work.
- #123: Reordering and renaming subplots after a split: This video explains how to order and rename the subplots after splitting the diagnostic plots according to covariate groups.
- #96: Comments : Good practice for project management and sharing: use “Comments” in the Monolix GUI to write down everything you consider important for your project.
- #127: mlxEditor – how to easily edit mlxtran models : Model editing is important in modeling and simulation workflows.This video shows how to do it in the mlxEditor – an application of MonolixSuite integrated with Monolix and Simulx and designed for the mlxtran language.
- #22 Preferences of Monolix : Would you like to export automatically all your plots, or keep a trace of all your runs? This kind of advanced features can be enabled in the Preferences.
- #55 How to set header preferences : Tired of tagging your data set columns? Learn how to define which columns headers you want to be automatically recognized in this video!
- #134 Display settings : Display settings of the MonolixSuite applications offer several options that might resolve display issues or come in useful or practical.
Input and output files
- #28 When to save, what to save to avoid losing information : Struggling reloading projects? Watch this video to understand the best saving workflow.
- #129 Dataset fingerprint : This video explains what is a fingerprint of a dataset and how it is used in Monolix and PKanalix to prevent the invalid results from loading.
Building the structural model
- #02 Using the built-in libraries of models : Discover our greatly enriched model libraries and how to efficiently browse through them.
- #09 Using the table statement : Learn about how to output variables such as the half-life or AUC from your monolix runs.
- #14 Writing a structural model with Mlxeditor : In Monolix, it is easy to write a new model or modify a model from the library without mistake thanks to the integrated text editor, with features like syntax highlighting and a compiling check.
- #16 Using regression variables in Monolix : This video shows with some examples how regressors can be used in Monolix.
- #20 Using the pkmodel() macro to define the PK part of a model : When writing a custom model, the pkmodel macro permits to define all typical PK models in one line of code.
- #26 Defining delay differential equations in MonolixSuite : Learn how to easily implement delay differential equations in the MonolixSuite for complex delay based PKPD models.
- #35 Introducing a scale factor to control parameter units : Wondering in which units your estimated parameters are? This video tells you how to change them using scale factors.
- #37 Understanding the error messages : Afraid of writing your own model? This video details the meaning of the most typical error m essages encountered when writing a new model.
- #38 Computing the AUC within a PK model : Computing the AUC can be done easily within a PK model. Watch this video to see an example of AUC simulation in Mlxplore, Monolix and Simulx.
- #39 Good practices for ODE-based models : Watch this video to get all the tricks to be sure that your ODE model will behave as you expect!
- #43 Understanding how and when the analytical solution is used : Wondering when the analytical solution of the model is used? This video explains all the details.
- #52 Using dose-related keywords in the structural model : The Mlxtran langage includes reserved keywords to use information from the dosing design in the structural model. Learn which ones in this video.
- #61 Mapping model outputs to data set observation types :The new mapping panel allows you to precisely map the data set observation ids to the model outputs and leave some out. Discover how in this video.
- #75 Focus on the depot macro : This video explains how to use the depot() macro to apply the doses defined in the data set to ODE variables in a model.
- #81 Alternative for DDEs : Did you know that many delayed differential equations (DDEs) can be rewritten as ODEs to improve the integration time? This video shows you how.
- #90: Calculating the NADIR or the Cmax in the structural model : The maximum or minimum of any ODE variable can be calculated directly in the structural model. This video shows you how to proceed.
- #93: Initial integration time : This video shows different initial conditions of a system of ODEs and their impact on the model.
- #130 Combining library models : In certain cases, a MonolixSuite user might want to combine different models from the Monolix library. This video explains how to do this by using the example of double absorption PK model combined with a PD model.
Building the statistical model
Random effects for individual parameters
- #06 Shrinkage : Shrinkage is said to bias diagnostic plots but Monolix has a special technique to get around the shrinkage problem. Watch the video to understand how it works.
- #31 Inter-individual variability with random effects : Which standard distribution should you choose for a parameter with random effect, and how can you verify that it is appropriate? Find out in this video.
- #32 Parameters with no variability : Find out how to remove random effect on a parameter, and how this affects its estimation.
- #33 Implementing a custom parameter distribution : Want to use another distribution than those implemented in Monolix? Learn how in this video.
- #08 Correlations between random effects : Two clicks are all it takes to define a correlation between random effects in Monolix 2018! As you will see in this video, and more.
- #63 Adapted logit-normal distribution : You know that your observations or individual parameters are bounded? Discover how to adapt the logit-normal distribution limits and increase the accuracy of your model.
- #64 Inter-occasion variability in Monolix : This video explains how several levels of variability can be combined in Monolix, such as inter-individual variability and inter-occasion variability.
Covariate effects for individual parameters
- #04 Transforming and adding covariates : This video unravels how to add covariates, and transform them to get the desired relationship.
- #15 Using time-varying covariates :Time-varying covariates can be used in Monolix, but they have to be defined in the structural model file. This video shows how.
- #27 Mixture of distributions with latent covariates : What is a latent covariate? How can it be used in Monolix to model mixtures of distributions for parameters? Find out in this video.
- #40 Defining a covariate-dependent standard deviation for a parameter : Find out how to define a parameter with different standard deviations for covariate groups
- #62 Calculating the typical value for each category of a categorical covariate :Are you perplexed by the beta parameters estimated by Monolix? This video shows you how to calculate the typical value for each category of a covariate.
- #70 Scaling of continuous covariates : Covariates are used to explain intra-individual variability of population parameters, but they can lower the confidence in the parameter estimation. Watch this video to see that a correct covariate scaling can prevent it.
- #135 Discretizing a continuous covariate into a categorical covariate : Continuous covariates can be discretized into categorical covariates using the Monolix covariate transformation panel in the GUI. This procedure allows to keep until all convenient covariate tools such as statistical tests.
- #05 Residual error models : How to diagnose and choose the best residual error model for continuous data in Monolix.
- #34 Using different error models for different studies : Find out how to model data from different studies with different error models while keeping the same population model.
Selection of statistical model
- #79 Understanding and using the statistical tests in Monolix :The statistical tests complement the diagnostic plots to guide the user in the development of the statistical model. This video explains what they mean and how to use them.
- #80 Proposal : Monolix is able to use the individual parameters of the current run to pre-test several covariate, correlation and error models. The most promising statistical model is displayed in the “proposal” section and can be applied in a single click, before running it in the population framework.
Tasks of Monolix and modeling workflow
- #07 Initial estimates : Choosing relevant initial estimates is crucial to get a fast convergence. Discover the features that guide this step in Monolix2018.
- #18 Tasks of the Monolix workflow : This video explains how the different estimation and diagnosis tasks in Monolix are used together in a workflow, and what their results are.
- #19 Using the “use last estimates” button : To speed up the convergence of a run, the initial parameter values can be read from the estimates of the previous run. Learn how and when to use it.
- #29 Exporting the plots data to replot elsewhere : Learn how to export the plots data from Monolix to reuse them with your other favorite software.
- #56 Reference in check initial estimates : Discover the new reference curve feature in the “check initial estimates” of Monolix in this video!
- #58 Automatic initialization of parameters : Setting up your PK model is now even faster, thanks to the new automatic initialization of parameters for models from the PK library.
- #69 Sycomore : We know that development of a good model is a long road of incremental improvements, which looks like a branched tree with different ideas and strategies. Sycomore is a visual and interactive tool designed to manage efficiently your Monolix runs and compare them side-by-side.
- #74 Files necessary to share a Monolix run or submit to the regulatory agencies : Want to include a Monolix run into a submission to regulatory agencies or share a run with somebody else? This video explains the files to include.
- #94: Individual parameters : This video shows how individual parameters are defined, estimated and where they are used in Monolix.
- #122: Generalized auto – initialization of parameters : This video describes the improvements done to the auto-init – a smart feature to initialize population parameters. There are some examples using models from the Monolix libraries and custom models, and it shows good practices in case of complex models.
Algorithms and results
Estimating population parameters
- #23 Interpreting the correlation matrix of the estimates : The correlation matrix can help detect a model overparameterization. Discover how in this video!
- #41 Calculating the coefficient of variation : The coefficient of variation can easily be calculated based on the Monolix outputs. This video shows you how.
- #42 Convergence assessment : Find out how easy it is to evaluate the convergence on multiple replicates.
- #44 Understanding how simulated annealing is used for parameter estimation : This video explains what is simulated annealing, an advanced setting of the parameter estimation algorithm, and in which case it may be disabled.
- #45: Understanding how SAEM works : Frustrated to use Monolix as a black box? This video explains what is exactly going on during the population parameters estimation.
- #46 A few useful SAEM settings : Not sure what the settings exactly mean? Discover the main ones in this video.
- #47 The convergence indicator : This video explains what the convergence indicator exactly is and how to use it.
- #49 Estimation methods and Bayesian approach : Bayesian estimation allows to take into account prior information for the estimation of parameters. Find out with this video how to use it in Monolix.
Estimating individual parameters
- #86 Estimating the conditional distributions : This video explains how the Monolix task “Conditional distribution” is computed and what is shown in the graphical report.
- #111: Demystifying probability distributions in Monolix : Several probability distributions are used in the Monolix algorithms. This video explains the meaning of each distribution and its closed form solution, when it exists.
- #113: How the EBEs are calculated : In this video, we explain the meaning of the EBEs and how they are calculated in Monolix.
- #48 What is the VPC, and how to get the most out of it : This video explains how the VPC is built and how to modify it in Monolix to get the most informative plot.
- #72 Display and informativeness of the BLQ data in the plots : To display BLQ data in the plot, Monolix uses simulated BLQ values. This video explains how these values are generated and why they improve the diagnostic power of the plots.
- #73 Typical patterns in the Obs versus Pred plot : Puzzled by the deviation you see in your Obs versus Pred plot? Learn how to interpret the most typical patterns!
- Mini case study: Advanced VPC: effect of censored data and dropout : If there are missing observations in your dataset, for example because of missing censored data or non-random dropout, chances are that the VPC is biased. This advanced understanding of the VPC is explained in this mini case-study, along with approaches to correct the bias with the MonolixSuite.
- #133 VPC with time after the last dose : A new feature of Monolix2021 allows to re-calculate the VPC plot with times after the last dose. It can make a more informative plot to diagnose the model and identify misspecifications.
Examples of specific models or data (continuous data only)
- #10 Working with several types of measurements : Find out how to match different types of observations to different model outputs, or use only part of your data set.
- #12 Modelling different types of drug administration : In Monolix, different types of administrations can be defined in the dataset to be flexibly modelled in different ways.
- #17 Taking censored data into account : Monolix automatically uses censored data information for estimation tasks, and simulates BLQ values for efficient graphical diagnosis of the model. All you have to do is tag the censored observations in the data set.
- #21 Defining additional doses with columns Steady State and Additional Doses : Multiple doses can be encoded in a compact way in a dataset handled by Monolix. Here is how it works.
- #25 Using the “ignored observation” column : Save time by working with a unique data set and selectively tagging columns as “ignored observation”
- #51 Encoding data with occasions : Occasions can be useful to define inter-occasion variability, or to use covariates that vary between occasions. There are several possible ways to encode occasions in a dataset to be used in the MonolixSuite. They are shown in this video.
- #53 Applying a washout or a selective reset : Do you want to reset your model? Watch this recap of the use of the EVENT ID column and reset() macro.
- #57 Writing a model for 2 drugs : You want to model two drugs at the same time? Watch this video to see how to define the mlxtran model.
- #66 Interpreting the PD versus PK plot – the example of hysteresis : The observation of the PK and PD data in Datxplore can bring useful insights to choose an appropriate model. Learn how to detect hysteresis in this video.
- #76 Calculate EBEs for a new data set using an existing model : You estimated a model on one data set and you want to use it for individual fits in another, for example sparse, data set? This video explains how to skip the re-estimation of population parameters when you load new data.
- #77 Generating predictive checks on an external data set : We show how to generate predictive checks, such as an external VPC, to check whether a population model estimated on a single dose study is also valid on a new multiple dose study for the same molecule.
- #83 Baseline PD models : The baseline of PD models can be estimated as a model parameter, fixed to the observed value or a mixture of both. This video shows you how to implement these three options.
- #85 Modeling exposure-response curves : This video shows how to use exposure-response models to obtain exposure-response curves directly in Monolix.
- #89 Missing data : This video shows different ways how Monolix treats datasets with missing information and what are the general guidelines to handle missing data.
- Mini case study: Time-varying clearance : In this mini-case study, we show step-by-step how to write phenomenological and mechanistic models in mlxtran language to capture the time-varying clearance of an example data set.
- Mini case study: Implementing a dose-dependent bioavailability : In this mini-case study, we show step-by-step how to implement models in mlxtran language to capture the dose-dependent bioavailability of an example data set.
- Mini case study: Writing a model for urine data : Did you know that Monolix can handle urine data? Discover how to format your data and write a urine model in this mini case study!
- Mini case study: Understanding flip-flop kinetics : Wondering what flip-flop kinetics really is and its consequences on NCA and popPK modeling? This mini case study explains it step-by-step with several examples.
- Mini case study: Calculating PK/PD parameters for an antibiotic : This case study shows the step-by-step modeling and simulation workflow for an antibiotic, in order to find the dosing regimen that optimizes the bacterial killing. PK/PD parameters such as Cmax/MIC, T above MIC and AUC/MIC will be calculated.
- #92: Splitting a dose into several fractions :This short video shows you the different ways of splitting a dose into 2, 3 or more fractions going via different routes. The same principles also apply to other fractions.
- #95: Developing a model for two formulations : Do you have a drug in two oral formulations in the same dataset? These formulations can differ in absorption and this video explains in detail how to model it.
Modeling non-continuous data
- #13 Encoding and representation of TTE data : Learn how to encode different types of time-to-event data for flexible modelling in Monolix.
- #50 Encoding and visualization of count and categorical data : Monolix also handles count and categorical data! Learn how to encode this type of data and how to explore it in Datxplore.
- #65 Time-to-event modeling with Monolix : Powerful modeling of time-to-event data with a parametric approach can be performed in Monolix, and is facilitated by a TTE models library. Discover how to do it in practice with this video.
- #82 Kaplan-Meier estimator : Survival function is a key function in the analysis of time-to-event data. In general, it is unknown and a typical way to estimate it is through the non-parametric Kaplan-Meier estimator. This video explains step by step how to construct it for exact and censored events and shows useful visualization features.
- #84 VPC for time-to-event data : VPC of survival data is a necessary diagnostic plot in the modeling of time-to-event data. This video explains how Monolix generates it for exact and interval censored events and presents settings that increase the modeling accuracy.
- #88 Mini case study: proportional hazard models : This video explains how proprtional hazard models can be implemented in Monolix.
- #30 Exploring a TMDD model with Mlxplore : The application Mlxplore of the MonolixSuite can help you identify the impact of some parameter on your model. This video shows how this is done with a TMDD model.
- #36 Exploring new dosing regimens in Mlxplore : Quick simulations of your model estimated with Monolix can be computed with Mlpxlore, for example to explore new dosing regimens.
- #97: Simulx interface : MonolixSuite2020 is now released! Here is a quick look at its main improvement: a new interface for Simulx, our applications for clinical trial simulations, and the new dark theme for MonolixSuite interface.
- #98:Importing a Monolix project to Simulx : Get started in Simulx by importing your favorite Monolix project. This video will show you how Simulx defines elements based on a monolix project to help you simulate new designs.
- #99:Defining a new treatment in Simulx : Simulx offers flexible options to define and simulate new dosing regimens. Discover them in this video!
- #101: Exploration in Simulx : With the exploration tab of Simulx, you can interactively explore new treatment designs and parameter values, and check the result on a typical prediction. Discover how with this video.
- #103: Simulation groups : Make the most of Simulx by using simulation groups! In this video, check how to simulate clinical trials with several arms.
- #104: Simulation outputs : This video shows the different options to define simulation outputs.
- #107: Easy sharing of your Simulx project : Have you noticed the option “save the user files in the results folder” in the Simulx project settings? This copies all the external files used in your project with the results, which is especially useful to move your simulations or share them with a colleague. Watch this video to know more!
- #108: Scaling a dose amount by a covariate in Simulx :This week, we have a deeper look at the option to scale dose amounts by covariates in Simulx. Why is this useful? How to choose the correct scaling? Here we provide three examples to help you get started with personalized treatments.
- #112: Additional lines in the model in Simulx : When you need a simulation output with a variable that is not defined in the Mlxtran model, then use the “additional lines in the model” feature to add new variables without modifying the original file. Watch this video for more details.
- #114: Using same individuals among groups in Simulx : When several simulation groups are simulated in Simulx, you have the choice to use same individuals among the groups. Check with this video how this feature works and when it is relevant.
- #115: Methods to sample from tables in Simulx : This video expains the different sampling methods available in Simulx for a simulation using an external table or tables created after an import from Monolix.
- #120: Sampling shared ids from external tables in Simulx : When Simulx samples individual values from several external tables, the option “shared ids” can be used to choose between independent and common samplings. Watch this video to see how it works and when it is useful.
- #124: Using the uncertainty of the population parameters in Simulx : This video explains how to take into account the uncertainty of the population parameters after importing a Monolix project.
- #126: Exporting simulated data as MonolixSuite formatted dataset : This video describes how to use the Export simulated data feature in Simulx and exportSimulatedData function in lixoftConnectors R package using an example of bioequivalence study design.
- #128: Sampling from the population parameter uncertainty in Simulx: behind the scene : This video explains the procedure Simulx uses to sample population parameters from their uncertainty distribution.
- #131: Applying uncertainty of the population parameters on the fixed effects only : This video shows how to take into account the uncertainty on the fixed effects but ignore the random effects in Simulx.
- #136 Bootstrap: visualization of uncertainty : Although parameter uncertainty can be visualized in Simulx using elements automatically created after importing a Monolix project, bootstrap estimates can be used to visualize uncertainty as well by saving estimates to a file and creating a population parameter element in Simulx that reads the parameters from this file.
- #138 Overlay Data on your Predictions : It is now possible to overlay data on your predictions in Simulx! Whether you are still exploring your data, whether you already have a good model or whether you are in the validation phase, we show you in this video how Simulx can guide you towards the next step with its new feature in the exploration tab.
- #54 Check lambda_z regression in PKanalix : For this first feature of the week on PKanalix, discover the “check lambda_z” tab, very useful to visualize and control the calculation of the terminal elimination phase λz.
- #59 Quick compartmental analysis with PKanalix :In addition to NCA, one of the main features of PKanalix is a quick calculation of PK parameters in the Compartmental Analysis framework.
- #67 Graphical results of NCA in PKanalix : Automatic plots are part of the efficient NCA workflow of PKanalix. Discover how to interpret and customize them!
- #102: Units in PKanalix : Now it is possible to manage and display NCA and CA results in your preferred metric units. It gives flexibility in reporting and makes the analysis more realistic. This video explains how to do it.
- #109: Selecting NCA parameters in PKanalix : The settings for non-compartmental analysis in PKanalix now include a convenient way to select a list of NCA parameters to compute and display in the results. The default list can also be customized.
- #110: Splitting the summary table in NCA results : In this video, discover a new feature of PKanalix that allows to split the non-compartmental analysis summary between different groups of individuals.
- #116: PKanalix behind the scenes: integral methods for AUC : This video, the first in a series dedicated to computation methods used in PKanalix, explains integral methods in the NCA task settings to compute AUC, AUMC and interpolation for partial AUC.
- #117: PKanalix behind the scenes: acceptance criteria : This video explains the NCA acceptance criteria used to flag profiles that satisfy specific conditions for the adjusted R2, for the percentage extrapolated AUC and for the SPAN.
- #118: PKanalix behind the scenes: lambda-Z : Lambda-z parameter corresponds to the slope of the terminal elimination phase. This video explains how linear regression, residuals and adjusted R2 coefficient are used to calculate it.
- #119: PKanalix behind the scenes: weighting : This video explains why we use the weighted objective function in the compartmental analysis in PKanalix and how different weights can account for data points with constant or proportional measurement noise.
- #132 Plotting the concentration profiles split by individual in PKanalix : This video shows how to plot pharmacokinetic profiles split by individual in PKanalix using the novel Individual fits plot.
- #68 Scripting Monolix in R : Sometimes you might want to use Monolix not through the interface but with scripts, for example to automate a set of actions in Monolix. The MonolixSuite comes with an API for R that allows to use Monolix and PKanalix from R, such that all you can do with the interface can be done with the API. Some examples are shown in this video.
- #71 Scripting PKanalix in R : Like Monolix, PKanalix has an API for R that can be used to easily automate non-compartmental and compartmental analyses and post-process the results. Watch some examples in this video!
- #125 Generating the diagnostic plots in R with lixoftConnectors : The lixoftConnectors R package, the MonolixSuite API for R, allows to script the MonolixSuite. Starting with the 2021 version of MonolixSuite, the plots of Monolix and PKanalix are available also as ggplot objects in R, thanks to new functions in the package lixoftConnectors.
- #136 Bootstrap : Bootstrapping is available for the Monolix projects through the bootmlx function of the Rsmlx R package.