Changelog
Source:NEWS.md
Version 0.3.2
- Add Kim 2006 model for IgG metabolism
- Add Xie 2019 agomelatine PK model
- Drop
qssince it will be archived
Version 0.3.1
CRAN release: 2025-10-22
- Bug fix for replacement of multiplicative expressions in
nlmixr2lib - phenylalanine_charbonneau_2021 had its net protein breakdown parameter corrected
- Kyhl_2016_nalmefene model was added
Version 0.3.0
CRAN release: 2024-10-07
Added ability to choose style type when modifying models. Currently supported styles are: “camel” for
variablesLikeThis, “snake” forvariables_like_this, “dot” forvariables.like.thisand “blank” forvariableslikethis. This can be selected withsetCombineType().With the new combination style, you can change how
etavariables are constructed with theoption(nlmixr2lib.etaCombineType="camel")or whatever you wish it to the variable style to be.-
Added new model building framework for building models
-
PK model building functions
addTransit()/removeTransit()which were present before, but now modified and made a bit more robust, more closely matching literature method of transit compartments.addDepot()/removeDepot()which were present before, but modified to be a bit more robust.addWeibullAbs()which adds a Weibull absorption to a PK modelconvertMM()converts linear elimination to Michaelis-Menten eliminationtransPK()converts theclstyle parameter transformations to various other PK transformations likek,aob,alpha,k12
PD model building functions
addIndirectLin()– this adds an indirect effect model to a PK model that has a concentrationCcin the model. This purposely uses a simple linear effect ofCc*EkorCc*Ikso it will be easy to parse and turn into other functional forms (likeEmaxorHill). If the PK model is not present it will useCcas a covariate in a purely PD models.addIndirect()– this builds onaddIndirectLin()and addsEmaxorHillmodels to a PK model. You can also setimax=1oremax=1to drop these parameters from being estimated in the model. Additionallyhill=TRUEwill add a Hill coefficient to the sigmoid model.addEffectCmtLin()– this adds an effect compartment based on theCcin the model. The linear effect can be modified into other function forms.addDirectLin()– this adds a direct effect model based on theCcin the model.-
Changing functional forms of Effect models
convertEmax()changes linear effect models to Emax modelsconvertEmaxHill()changes linear effect models to Hill modelsconvertQuad()changes linear effect models to quadratic modelsconvertLogLin()changes linear effect models to log-linear models
-
Changing functional forms of Baselines in non-indirect response models
addBaselineConst()changes the zero baseline to a estimated constantaddBaselineLin()changes the zero baseline to a estimated constant and a linear constant with respect totime.addBaselineExp()changes the zero baseline to a exponential decay with respect to timeaddBaseline1exp()– the baseline effect is changed from zero to to an exponential approaching to a constant (with respect to time).
-
Changing model properties (all use
addCmtProp())addBioavailability()adds bioavailability property to a compartmentaddRate()adds a modeled rate to a compartmentaddDur()adds modeled duration to a compartmentaddIni()adds an initial value to a compartmentaddLag()adds a lag time to the a compartment
-
Add Carlsson Petri (2021) liraglutide PK model
Add Cirincione (2017) exenatide immediate-release PK model
Add a variety of indirect response models
Add a variety of tumor growth inhibition models and move all oncology models into a new model database directory
Add a variety of double-absorption PK models
cpand relatedcpddSdandcppropSdwere renamed toCc,CcAddSdandCcPropSd(fix #70).Multiple-endpoint models will have the
DVcolumn in the modeldb separated by commas.
Version 0.2.0
CRAN release: 2023-03-29
- Work with the new
rxode2version 2.0.12model()andini()assignment methods. - Therapeutic-area specific models have begun being added.
- Models can now give the user some additional information load via the
messagemeta-data. - Models can now be in different directories. The change is for ease of maintaining the library, it is not a change that affects users.
- A regression where
addEta()did not change the parameter, related to a change inrxode2, was fixed. -
addEta()detects where to add etas more robustly when covariates are on the parameter.