Skip to contents

This functions sets up a nonlinear mixed effects model for an mmkin row object. An mmkin row object is essentially a list of mkinfit objects that have been obtained by fitting the same model to a list of datasets.

Usage

# S3 method for mmkin
nlme(
  model,
  data = "auto",
  fixed = lapply(as.list(names(mean_degparms(model))), function(el) eval(parse(text =
    paste(el, 1, sep = "~")))),
  random = pdDiag(fixed),
  groups,
  start = mean_degparms(model, random = TRUE, test_log_parms = TRUE),
  correlation = NULL,
  weights = NULL,
  subset,
  method = c("ML", "REML"),
  na.action = na.fail,
  naPattern,
  control = list(),
  verbose = FALSE
)

# S3 method for nlme.mmkin
print(x, digits = max(3, getOption("digits") - 3), ...)

# S3 method for nlme.mmkin
update(object, ...)

Arguments

model

An mmkin row object.

data

Ignored, data are taken from the mmkin model

fixed

Ignored, all degradation parameters fitted in the mmkin model are used as fixed parameters

random

If not specified, no correlations between random effects are set up for the optimised degradation model parameters. This is achieved by using the nlme::pdDiag method.

groups

See the documentation of nlme

start

If not specified, mean values of the fitted degradation parameters taken from the mmkin object are used

correlation

See the documentation of nlme

weights

passed to nlme

subset

passed to nlme

method

passed to nlme

na.action

passed to nlme

naPattern

passed to nlme

control

passed to nlme

verbose

passed to nlme

x

An nlme.mmkin object to print

digits

Number of digits to use for printing

...

Update specifications passed to update.nlme

object

An nlme.mmkin object to update

Value

Upon success, a fitted 'nlme.mmkin' object, which is an nlme object with additional elements. It also inherits from 'mixed.mmkin'.

Details

Note that the convergence of the nlme algorithms depends on the quality of the data. In degradation kinetics, we often only have few datasets (e.g. data for few soils) and complicated degradation models, which may make it impossible to obtain convergence with nlme.

Note

As the object inherits from nlme::nlme, there is a wealth of methods that will automatically work on 'nlme.mmkin' objects, such as nlme::intervals(), nlme::anova.lme() and nlme::coef.lme().

Examples

ds <- lapply(experimental_data_for_UBA_2019[6:10],
 function(x) subset(x$data[c("name", "time", "value")], name == "parent"))

# \dontrun{
  f <- mmkin(c("SFO", "DFOP"), ds, quiet = TRUE, cores = 1)
  library(nlme)
  f_nlme_sfo <- nlme(f["SFO", ])
  f_nlme_dfop <- nlme(f["DFOP", ])
  anova(f_nlme_sfo, f_nlme_dfop)
#>             Model df      AIC      BIC    logLik   Test  L.Ratio p-value
#> f_nlme_sfo      1  5 625.0539 637.5529 -307.5269                        
#> f_nlme_dfop     2  9 495.1270 517.6253 -238.5635 1 vs 2 137.9269  <.0001
  print(f_nlme_dfop)
#> Kinetic nonlinear mixed-effects model fit by maximum likelihood
#> 
#> Structural model:
#> d_parent/dt = - ((k1 * g * exp(-k1 * time) + k2 * (1 - g) * exp(-k2 *
#>            time)) / (g * exp(-k1 * time) + (1 - g) * exp(-k2 * time)))
#>            * parent
#> 
#> Data:
#> 90 observations of 1 variable(s) grouped in 5 datasets
#> 
#> Log-likelihood: -238.6
#> 
#> Fixed effects:
#>  list(parent_0 ~ 1, log_k1 ~ 1, log_k2 ~ 1, g_qlogis ~ 1) 
#> parent_0   log_k1   log_k2 g_qlogis 
#>  94.1702  -1.8002  -4.1474   0.0324 
#> 
#> Random effects:
#>  Formula: list(parent_0 ~ 1, log_k1 ~ 1, log_k2 ~ 1, g_qlogis ~ 1)
#>  Level: ds
#>  Structure: Diagonal
#>         parent_0 log_k1 log_k2 g_qlogis Residual
#> StdDev:    2.488 0.8447   1.33   0.4652    2.321
#> 
  plot(f_nlme_dfop)

  endpoints(f_nlme_dfop)
#> $distimes
#>            DT50     DT90 DT50back  DT50_k1  DT50_k2
#> parent 10.79857 100.7937 30.34192 4.193937 43.85442
#> 

  ds_2 <- lapply(experimental_data_for_UBA_2019[6:10],
   function(x) x$data[c("name", "time", "value")])
  m_sfo_sfo <- mkinmod(parent = mkinsub("SFO", "A1"),
    A1 = mkinsub("SFO"), use_of_ff = "min", quiet = TRUE)
  m_sfo_sfo_ff <- mkinmod(parent = mkinsub("SFO", "A1"),
    A1 = mkinsub("SFO"), use_of_ff = "max", quiet = TRUE)
  m_dfop_sfo <- mkinmod(parent = mkinsub("DFOP", "A1"),
    A1 = mkinsub("SFO"), quiet = TRUE)

  f_2 <- mmkin(list("SFO-SFO" = m_sfo_sfo,
   "SFO-SFO-ff" = m_sfo_sfo_ff,
   "DFOP-SFO" = m_dfop_sfo),
    ds_2, quiet = TRUE)

  f_nlme_sfo_sfo <- nlme(f_2["SFO-SFO", ])
  plot(f_nlme_sfo_sfo)


  # With formation fractions this does not coverge with defaults
  # f_nlme_sfo_sfo_ff <- nlme(f_2["SFO-SFO-ff", ])
  #plot(f_nlme_sfo_sfo_ff)

  # For the following, we need to increase pnlsMaxIter and the tolerance
  # to get convergence
  f_nlme_dfop_sfo <- nlme(f_2["DFOP-SFO", ],
    control = list(pnlsMaxIter = 120, tolerance = 5e-4))

  plot(f_nlme_dfop_sfo)


  anova(f_nlme_dfop_sfo, f_nlme_sfo_sfo)
#>                 Model df       AIC      BIC    logLik   Test  L.Ratio p-value
#> f_nlme_dfop_sfo     1 13  843.8547  884.620 -408.9273                        
#> f_nlme_sfo_sfo      2  9 1085.1821 1113.404 -533.5910 1 vs 2 249.3274  <.0001

  endpoints(f_nlme_sfo_sfo)
#> $ff
#> parent_sink   parent_A1     A1_sink 
#>   0.5912432   0.4087568   1.0000000 
#> 
#> $distimes
#>            DT50     DT90
#> parent 19.13518  63.5657
#> A1     66.02155 219.3189
#> 
  endpoints(f_nlme_dfop_sfo)
#> $ff
#>   parent_A1 parent_sink 
#>   0.2768574   0.7231426 
#> 
#> $distimes
#>             DT50     DT90 DT50back  DT50_k1  DT50_k2
#> parent  11.07091 104.6320 31.49737 4.462383 46.20825
#> A1     162.30550 539.1672       NA       NA       NA
#> 

  if (length(findFunction("varConstProp")) > 0) { # tc error model for nlme available
    # Attempts to fit metabolite kinetics with the tc error model are possible,
    # but need tweeking of control values and sometimes do not converge

    f_tc <- mmkin(c("SFO", "DFOP"), ds, quiet = TRUE, error_model = "tc")
    f_nlme_sfo_tc <- nlme(f_tc["SFO", ])
    f_nlme_dfop_tc <- nlme(f_tc["DFOP", ])
    AIC(f_nlme_sfo, f_nlme_sfo_tc, f_nlme_dfop, f_nlme_dfop_tc)
    print(f_nlme_dfop_tc)
  }
#> Kinetic nonlinear mixed-effects model fit by maximum likelihood
#> 
#> Structural model:
#> d_parent/dt = - ((k1 * g * exp(-k1 * time) + k2 * (1 - g) * exp(-k2 *
#>            time)) / (g * exp(-k1 * time) + (1 - g) * exp(-k2 * time)))
#>            * parent
#> 
#> Data:
#> 90 observations of 1 variable(s) grouped in 5 datasets
#> 
#> Log-likelihood: -238.4
#> 
#> Fixed effects:
#>  list(parent_0 ~ 1, log_k1 ~ 1, log_k2 ~ 1, g_qlogis ~ 1) 
#> parent_0   log_k1   log_k2 g_qlogis 
#> 94.04774 -1.82340 -4.16716  0.05685 
#> 
#> Random effects:
#>  Formula: list(parent_0 ~ 1, log_k1 ~ 1, log_k2 ~ 1, g_qlogis ~ 1)
#>  Level: ds
#>  Structure: Diagonal
#>         parent_0 log_k1 log_k2 g_qlogis Residual
#> StdDev:    2.474   0.85  1.337   0.4659        1
#> 
#> Variance function:
#>  Structure: Constant plus proportion of variance covariate
#>  Formula: ~fitted(.) 
#>  Parameter estimates:
#>      const       prop 
#> 2.23222933 0.01262399 

  f_2_obs <- update(f_2, error_model = "obs")
  f_nlme_sfo_sfo_obs <- nlme(f_2_obs["SFO-SFO", ])
  print(f_nlme_sfo_sfo_obs)
#> Kinetic nonlinear mixed-effects model fit by maximum likelihood
#> 
#> Structural model:
#> d_parent/dt = - k_parent_sink * parent - k_parent_A1 * parent
#> d_A1/dt = + k_parent_A1 * parent - k_A1_sink * A1
#> 
#> Data:
#> 170 observations of 2 variable(s) grouped in 5 datasets
#> 
#> Log-likelihood: -473
#> 
#> Fixed effects:
#>  list(parent_0 ~ 1, log_k_parent_sink ~ 1, log_k_parent_A1 ~ 1,      log_k_A1_sink ~ 1) 
#>          parent_0 log_k_parent_sink   log_k_parent_A1     log_k_A1_sink 
#>            87.976            -3.670            -4.164            -4.645 
#> 
#> Random effects:
#>  Formula: list(parent_0 ~ 1, log_k_parent_sink ~ 1, log_k_parent_A1 ~ 1,      log_k_A1_sink ~ 1)
#>  Level: ds
#>  Structure: Diagonal
#>         parent_0 log_k_parent_sink log_k_parent_A1 log_k_A1_sink Residual
#> StdDev:    3.992             1.777           1.055        0.4821    6.483
#> 
#> Variance function:
#>  Structure: Different standard deviations per stratum
#>  Formula: ~1 | name 
#>  Parameter estimates:
#>    parent        A1 
#> 1.0000000 0.2050005 
  f_nlme_dfop_sfo_obs <- nlme(f_2_obs["DFOP-SFO", ],
    control = list(pnlsMaxIter = 120, tolerance = 5e-4))

  f_2_tc <- update(f_2, error_model = "tc")
  # f_nlme_sfo_sfo_tc <- nlme(f_2_tc["SFO-SFO", ]) # No convergence with 50 iterations
  # f_nlme_dfop_sfo_tc <- nlme(f_2_tc["DFOP-SFO", ],
  #  control = list(pnlsMaxIter = 120, tolerance = 5e-4)) # Error in X[, fmap[[nm]]] <- gradnm

  anova(f_nlme_dfop_sfo, f_nlme_dfop_sfo_obs)
#>                     Model df      AIC     BIC    logLik   Test  L.Ratio p-value
#> f_nlme_dfop_sfo         1 13 843.8547 884.620 -408.9273                        
#> f_nlme_dfop_sfo_obs     2 14 817.5338 861.435 -394.7669 1 vs 2 28.32084  <.0001

# }