Lixoft monolix Suite 2024R1
Lixoft develops modeling and simulation software for advanced model-based drug development. Its products provide powerful and easy solutions for population analysis in preclinical, clinical trials and treatment personalization.
Lixoft’s technology is the result of a decade of statistical and modeling research projects, led by Inria, for advanced population analysis, nonlinear mixed effects models in pharmacology, preclinical and clinical trial modeling and simulation. Monolix is the most advanced and simple solution for nonlinear mixed effects modeling (NLME) in pharmacology.
It is based on the SAEM algorithm and provides strong global convergence even for complex PK/PD models. Monolix is used for preclinical and clinical population PK/PD modeling and systems pharmacology.
Lixoft Monolix Suite generates Visual Predictive Check instantly, generating complete diagnostic charts for your complex PK/PD models. It covers a wide range of data types as well as statistical features for population PK/PD modeling.
New Features
1. Global Interoperability
* Extended interoperability to export PKanalix, Monolix or Simulx runs to any other application of MonolixSuite.
In addition to the existing export/import, it is now possible to:
– Export a PKanalix CA run to Simulx to predict the response of these same individuals to another dosing regimen.
– Export a Simulx run or a Monolix individual fit or VPC simulation to PKanalix to run an NCA on the simulated data.
2. Global Display
* Report module for automatic report generation based on custom Word template files:
– Placeholders can be written to any Word template file, and when the report is generated, they will be replaced by plots and tables from the PKanalix or Monolix run
– Placeholders can be easily generated for each plot and table in the GUI and then copied into the Word template
– Placeholders for plots and tables can be customized to display what you want in the report
* Custom axis limits can be set for observed data and VPC plots.
* In the “Observed Data” plot, trend lines (mean and standard error) for different groups can be merged on the same plot after splitting by covariates.
* Software feedback can now be sent directly from the interface of any application of MonolixSuite.
3. Datasets
* Data formatting module in the interface to make input format more flexible.
From input datasets (not yet in MonolixSuite format), it is now possible to:
– Merge multiple observation types
– Add censoring information based on BLQ labels
– Add treatment information from an external file, or manually specify multiple possibilities (multiple dose regimens, repeated cycles, infusions, combination therapies, dose depends on class, use dose intervals, etc.)
– Add covariate columns and regression columns from another file
4. PKanalix
* [Units] Quantity columns, other standardized units per body weight or body surface area.
* [CA Tasks] Any model in CA can be used, written by yourself using mlxtran syntax or from MonolixSuite libraries.
* [CA Tasks] Improved algorithms in CA to escape better local minima and improve fit (important for models more complex than standard pk).
* [CA Tasks] New cost function based on generalized least squares, with the ability to compare models with AIC and BIC.
*[CA Task] New weighting option 1/|Ypred*Yobs|.
*[CA Results] Individual costs are reported in the individual parameter tables, and a new Costs tab shows values for global costs, -2LL, AIC, and BIC.
*[CA Plots] For CA, new observed vs. predicted plots are available to diagnose fit.
*[Results] Additional summary measures Geometric CV and Harmonic Mean.
*[Preferences] Custom aliases for NCA parameter names and summary measure names, which appear in tables, plots, and reports.
5. Monolix
*[Results] Coefficient of variation is now reported in addition to standard deviation for random effects in the population parameter table.
*[Data] Sequential methods are now easier, as PK parameters from another run can be added to a dataset as a regression function directly in the interface (via the Data Formatting module).
6. Simulx
*[Results] In tables and plots, map simulation IDs of individuals to original identifiers from a Monolix run or user-defined external elements.
*[Plots] Distributions of simulation output from different groups can now be merged on the same plot.
*[Definition] Import new elements created after Monolix run to simulate the effect of covariation or parameter uncertainty on a typical individual (omegas set to zero).
*[Results] More post-processing possibilities for calculation results, such as
– duration below, above or between specific values,
– output values for custom times,
– normalization relative to min/max values (in addition to baseline values)
– define results for reporting time as continuous or event type
*[Definition] Convert single elements to population elements and vice versa with one click.
*[Plot] Visual cues automatically switch to OFF after setting them.
7. Lixoft Connector/Command Line
* New Monolix connector for drawing from R.
* New Simulx connector for defining and calculating results and endpoints in the interface.
* New connector covers new functions available in the interface (formatData, generateReport, getCACost).
* New function reportGenerator to generate reports directly from the command line.
* New connector for selecting library models using the same filters as in the GUI.
* Improved error and warning messages when parameters are not recognized.
* Extensive documentation for the Simulx connector, including detailed descriptions of each function and working examples.
* setConsoleMode (show calculation progress in the GUI) now also works for Simulx and PKanalix.
8. Updates:
* ODE solver CVODE upgraded to version 5.5.
9. Fixed bugs:
===== Global interoperability
* [export] Fixed slow file export when exporting on a shared drive.
===== Global display
* [observed data] When removing points and lines, dosing times set to always visible disappear.
* [observed data] Axis limits are based on points, even if the points are not displayed on the plot.
* [plots] For all plots, the “random colors” option does not work (nothing happens when you click it).
* [observed data] When a plot is split and paged, the legend is not updated correctly when going from one page to another.
* [libraries] The parameters displayed when selecting a model from a library are not always the input parameters of the model.
===== datasets
*[data explanation] In some very specific cases, when a dataset with “1” in the CENS column is loaded, it is considered right-censored instead of left-censored.
*[covariate statistics] Stratified covariates do not appear in the covariate statistics tab.
====PKanalix
*[compute] Acceptance criteria are used to filter individuals even for parameters that do not use lambda_z.
*[settings] If partial AUC is selected, it is not possible to remove any partial parameter from the list of computed parameters.
*[results] In the individual evaluation table, ID and occasion are displayed as doubles instead of integers.
*[data]PKanalix crashes when reloading an item in a dataset that uses Cc as a header.
*[preferences] Significant figures are not applied to all values in the results table (only to the currently displayed row).
===== monolith
*[check initial estimates] Observation models other than normal distribution are not handled correctly by the automatic initialization algorithm.
*[Statistical Models] The names of the wrong parameters were based on the default observation names given by Monolix first after loading the model, which were not modified if the mapping and observation names were changed manually.
*[Statistical Models] The vertical scroll bar disappeared when showing formulas.
*[plots/VPC] Selecting and deselecting “Time since last dose” did not return the box to the same state.
===== SycMo Figs
* In the Comparison tab, NaNs for RSE were not displayed, and the cells were displayed as empty.
* When selecting a folder containing only one mlxtran file, the folder and “Scan” buttons did not appear.
===== Simulations
*[Simulations] Variability between time periods was not accounted for when there was only one time period.
*[Results] Endpoint summary tabs were displayed in Results even when there was only one replicate.
*[Results] Renamed results were duplicated with results in the group comparison summary.
===== Lixoft Connector/Command Line
*[getChartsData] getChartsData computes chart data for all plots, even when a specific plot was requested.
*[defineTreatmentElement] Option to scale treatment by covariate is available in GUI but not in connector.
*[definePopulationElement] Crash instead of error when trying to set population parameter element with multiple lines.
*[setGroupElement] Error when setting multiple elements to the same simulation group.
*[command line] Options –gui are not taken into account if option “-t monolix” is used.
*[newProject]newProject cannot use modelFile=”lib:…”.
*[deleteOccasionElement]deleteOccasionElement deletes mlx_Cov if no occasions exist in the project.
*[setData]setData (and Rsmlx functions that use it: bootmlx, confintmlx) did not correctly set data on some machines in some cases, resulting in error “Observation name inconsistency between observation type and data file input”.
*[define..Element] Occasion definition of common element did not properly take into account the occasion column.
*[getTreatmentElements]getTreatmentElements() returned a truncated number of rows.