path: root/R/mkinmod.R

#' Function to set up a kinetic model with one or more state variables
#'
#' The function usually takes several expressions, each assigning a compound
#' name to a list, specifying the kinetic model type and reaction or transfer
#' to other observed compartments. Instead of specifying several expressions, a
#' list of lists can be given in the speclist argument.
#'
#' For the definition of model types and their parameters, the equations given
#' in the FOCUS and NAFTA guidance documents are used.
#'
#' @param ...  For each observed variable, a list has to be specified as an
#'   argument, containing at least a component \code{type}, specifying the type
#'   of kinetics to use for the variable. Currently, single first order
#'   kinetics "SFO", indeterminate order rate equation kinetics "IORE", or
#'   single first order with reversible binding "SFORB" are implemented for all
#'   variables, while "FOMC", "DFOP" and "HS" can additionally be chosen for
#'   the first variable which is assumed to be the source compartment.
#'   Additionally, each component of the list can include a character vector
#'   \code{to}, specifying names of variables to which a transfer is to be
#'   assumed in the model.  If the argument \code{use_of_ff} is set to "min"
#'   (default) and the model for the compartment is "SFO" or "SFORB", an
#'   additional component of the list can be "sink=FALSE" effectively fixing
#'   the flux to sink to zero.
#' @param use_of_ff Specification of the use of formation fractions in the
#'   model equations and, if applicable, the coefficient matrix. If "min", a
#'   minimum use of formation fractions is made in order to avoid fitting the
#'   product of formation fractions and rate constants. If "max", formation
#'   fractions are always used.
#' @param speclist The specification of the observed variables and their
#'   submodel types and pathways can be given as a single list using this
#'   argument. Default is NULL.
#' @param quiet Should messages be suppressed?
#' @param verbose If \code{TRUE}, passed to \code{\link{cfunction}} if
#'   applicable to give detailed information about the C function being built.
#' @importFrom methods signature
#' @importFrom pkgbuild has_compiler
#' @importFrom inline cfunction
#' @return A list of class \code{mkinmod} for use with \code{\link{mkinfit}},
#'   containing, among others,
#'   \item{diffs}{
#'     A vector of string representations of differential equations, one for
#'     each modelling variable.
#'   }
#'   \item{map}{
#'     A list containing named character vectors for each observed variable,
#'     specifying the modelling variables by which it is represented.
#'   }
#'   \item{use_of_ff}{
#'     The content of \code{use_of_ff} is passed on in this list component.
#'   }
#'   \item{coefmat}{
#'     The coefficient matrix, if the system of differential equations can be
#'     represented by one.
#'   }
#'   \item{cf}{
#'     If generated, the compiled function as returned by cfunction.
#'   }
#' @note The IORE submodel is not well tested for metabolites. When using this
#'   model for metabolites, you may want to read the second note in the help
#'   page to \code{\link{mkinfit}}.
#' @author Johannes Ranke
#' @references FOCUS (2006) \dQuote{Guidance Document on Estimating Persistence
#'   and Degradation Kinetics from Environmental Fate Studies on Pesticides in
#'   EU Registration} Report of the FOCUS Work Group on Degradation Kinetics,
#'   EC Document Reference Sanco/10058/2005 version 2.0, 434 pp,
#'   \url{http://esdac.jrc.ec.europa.eu/projects/degradation-kinetics}
#'
#'   NAFTA Technical Working Group on Pesticides (not dated) Guidance for
#'   Evaluating and Calculating Degradation Kinetics in Environmental Media
#' @examples
#'
#' # Specify the SFO model (this is not needed any more, as we can now mkinfit("SFO", ...)
#' SFO <- mkinmod(parent = list(type = "SFO"))
#'
#' # One parent compound, one metabolite, both single first order
#' SFO_SFO <- mkinmod(
#'   parent = mkinsub("SFO", "m1"),
#'   m1 = mkinsub("SFO"))
#'
#' \dontrun{
#' # The above model used to be specified like this, before the advent of mkinsub()
#' SFO_SFO <- mkinmod(
#'   parent = list(type = "SFO", to = "m1"),
#'   m1 = list(type = "SFO"))
#'
#' # Show details of creating the C function
#' SFO_SFO <- mkinmod(
#'   parent = mkinsub("SFO", "m1"),
#'   m1 = mkinsub("SFO"), verbose = TRUE)
#'
#' # If we have several parallel metabolites
#' # (compare tests/testthat/test_synthetic_data_for_UBA_2014.R)
#' m_synth_DFOP_par <- mkinmod(parent = mkinsub("DFOP", c("M1", "M2")),
#'                            M1 = mkinsub("SFO"),
#'                            M2 = mkinsub("SFO"),
#'                            use_of_ff = "max", quiet = TRUE)
#'
#' fit_DFOP_par_c <- mkinfit(m_synth_DFOP_par,
#'                           synthetic_data_for_UBA_2014[[12]]$data,
#'                           quiet = TRUE)
#' }
#'
#' @export mkinmod
mkinmod <- function(..., use_of_ff = "min", speclist = NULL, quiet = FALSE, verbose = FALSE)
{
  if (is.null(speclist)) spec <- list(...)
  else spec <- speclist
  obs_vars <- names(spec)

  # Check if any of the names of the observed variables contains any other
  for (obs_var in obs_vars) {
    if (length(grep(obs_var, obs_vars)) > 1) stop("Sorry, variable names can not contain each other")
    if (grepl("_to_", obs_var)) stop("Sorry, names of observed variables can not contain _to_")
    if (obs_var == "sink") stop("Naming a compound 'sink' is not supported")
  }

  if (!use_of_ff %in% c("min", "max"))
    stop("The use of formation fractions 'use_of_ff' can only be 'min' or 'max'")

  # The returned model will be a list of character vectors, containing {{{
  # differential equations (if supported), parameter names and a mapping from
  # model variables to observed variables. If possible, a matrix representation
  # of the differential equations is included
  # Compiling the functions from the C code generated below only works if the
  # implicit assumption about differential equations specified below
  # is satisfied
  parms <- vector()
  # }}}

  # Do not return a coefficient matrix mat when FOMC, IORE, DFOP, HS or logistic is used for the parent {{{
  if(spec[[1]]$type %in% c("FOMC", "IORE", "DFOP", "HS", "logistic")) {
    mat = FALSE
  } else mat = TRUE
  #}}}

  # Establish a list of differential equations as well as a map from observed {{{
  # compartments to differential equations
  diffs <- vector()
  map <- list()
  for (varname in obs_vars)
  {
    # Check the type component of the compartment specification {{{
    if(is.null(spec[[varname]]$type)) stop(
      "Every part of the model specification must be a list containing a type component")
    if(!spec[[varname]]$type %in% c("SFO", "FOMC", "IORE", "DFOP", "HS", "SFORB", "logistic")) stop(
      "Available types are SFO, FOMC, IORE, DFOP, HS, SFORB and logistic only")
    if(spec[[varname]]$type %in% c("FOMC", "DFOP", "HS", "logistic") & match(varname, obs_vars) != 1) {
        stop(paste("Types FOMC, DFOP, HS and logistic are only implemented for the first compartment,",
                   "which is assumed to be the source compartment"))
    }
    #}}}
    # New (sub)compartments (boxes) needed for the model type {{{
    new_boxes <- switch(spec[[varname]]$type,
      SFO = varname,
      FOMC = varname,
      IORE = varname,
      DFOP = varname,
      HS = varname,
      logistic = varname,
      SFORB = paste(varname, c("free", "bound"), sep = "_")
    )
    map[[varname]] <- new_boxes
    names(map[[varname]]) <- rep(spec[[varname]]$type, length(new_boxes)) #}}}
    # Start a new differential equation for each new box {{{
    new_diffs <- paste("d_", new_boxes, " =", sep = "")
    names(new_diffs) <- new_boxes
    diffs <- c(diffs, new_diffs) #}}}
  } #}}}

  # Create content of differential equations and build parameter list {{{
  for (varname in obs_vars)
  {
    # Get the name of the box(es) we are working on for the decline term(s)
    box_1 = map[[varname]][[1]] # This is the only box unless type is SFORB
    # Turn on sink if this is not explicitly excluded by the user by
    # specifying sink=FALSE
    if(is.null(spec[[varname]]$sink)) spec[[varname]]$sink <- TRUE
    if(spec[[varname]]$type %in% c("SFO", "IORE", "SFORB")) { # {{{ Add decline term
      if (use_of_ff == "min") { # Minimum use of formation fractions
        if(spec[[varname]]$type == "IORE" && length(spec[[varname]]$to) > 0) {
           stop("Transformation reactions from compounds modelled with IORE\n",
                "are only supported with formation fractions (use_of_ff = 'max')")
        }
        if(spec[[varname]]$sink) {
          # If sink is requested, add first-order/IORE sink term
          k_compound_sink <- paste("k", box_1, "sink", sep = "_")
          if(spec[[varname]]$type == "IORE") {
            k_compound_sink <- paste("k__iore", box_1, "sink", sep = "_")
          }
          parms <- c(parms, k_compound_sink)
          decline_term <- paste(k_compound_sink, "*", box_1)
          if(spec[[varname]]$type == "IORE") {
            N <- paste("N", box_1, sep = "_")
            parms <- c(parms, N)
            decline_term <- paste0(decline_term, "^", N)
          }
        } else { # otherwise no decline term needed here
          decline_term = "0"
        }
      } else { # Maximum use of formation fractions
        k_compound <- paste("k", box_1, sep = "_")
        if(spec[[varname]]$type == "IORE") {
          k_compound <- paste("k__iore", box_1, sep = "_")
        }
        parms <- c(parms, k_compound)
        decline_term <- paste(k_compound, "*", box_1)
        if(spec[[varname]]$type == "IORE") {
          N <- paste("N", box_1, sep = "_")
          parms <- c(parms, N)
          decline_term <- paste0(decline_term, "^", N)
        }
      }
    } #}}}
    if(spec[[varname]]$type == "FOMC") { # {{{ Add FOMC decline term
      # From p. 53 of the FOCUS kinetics report, without the power function so it works in C
      decline_term <- paste("(alpha/beta) * 1/((time/beta) + 1) *", box_1)
      parms <- c(parms, "alpha", "beta")
    } #}}}
    if(spec[[varname]]$type == "DFOP") { # {{{ Add DFOP decline term
      # From p. 57 of the FOCUS kinetics report
      decline_term <- paste("((k1 * g * exp(-k1 * time) + k2 * (1 - g) * exp(-k2 * time)) / (g * exp(-k1 * time) + (1 - g) * exp(-k2 * time))) *", box_1)
      parms <- c(parms, "k1", "k2", "g")
    } #}}}
    HS_decline <- "ifelse(time <= tb, k1, k2)" # Used below for automatic translation to C
    if(spec[[varname]]$type == "HS") { # {{{ Add HS decline term
      # From p. 55 of the FOCUS kinetics report
      decline_term <- paste(HS_decline, "*", box_1)
      parms <- c(parms, "k1", "k2", "tb")
    } #}}}
    if(spec[[varname]]$type == "logistic") { # {{{ Add logistic decline term
      # From p. 67 of the FOCUS kinetics report (2014)
      decline_term <- paste("(k0 * kmax)/(k0 + (kmax - k0) * exp(-r * time)) *", box_1)
      parms <- c(parms, "kmax", "k0", "r")
    } #}}}
    # Add origin decline term to box 1 (usually the only box, unless type is SFORB)#{{{
    diffs[[box_1]] <- paste(diffs[[box_1]], "-", decline_term)#}}}
    if(spec[[varname]]$type == "SFORB") { # {{{ Add SFORB reversible binding terms
      box_2 = map[[varname]][[2]]
      k_free_bound <- paste("k", varname, "free", "bound", sep = "_")
      k_bound_free <- paste("k", varname, "bound", "free", sep = "_")
      parms <- c(parms, k_free_bound, k_bound_free)
      reversible_binding_term_1 <- paste("-", k_free_bound, "*", box_1, "+",
        k_bound_free, "*", box_2)
      reversible_binding_term_2 <- paste("+", k_free_bound, "*", box_1, "-",
        k_bound_free, "*", box_2)
      diffs[[box_1]] <- paste(diffs[[box_1]], reversible_binding_term_1)
      diffs[[box_2]] <- paste(diffs[[box_2]], reversible_binding_term_2)
    } #}}}

    # Transfer between compartments#{{{
    to <- spec[[varname]]$to
    if(!is.null(to)) {
      # Name of box from which transfer takes place
      origin_box <- box_1

      # Number of targets
      n_targets = length(to)

      # Add transfer terms to listed compartments
      for (target in to) {
        if (!target %in% obs_vars) stop("You did not specify a submodel for target variable ", target)
        target_box <- switch(spec[[target]]$type,
          SFO = target,
          IORE = target,
          SFORB = paste(target, "free", sep = "_"))
        if (use_of_ff == "min" && spec[[varname]]$type %in% c("SFO", "SFORB"))
        {
          k_from_to <- paste("k", origin_box, target_box, sep = "_")
          parms <- c(parms, k_from_to)
          diffs[[origin_box]] <- paste(diffs[[origin_box]], "-",
            k_from_to, "*", origin_box)
          diffs[[target_box]] <- paste(diffs[[target_box]], "+",
            k_from_to, "*", origin_box)
        } else {
          # Do not introduce a formation fraction if this is the only target
          if (spec[[varname]]$sink == FALSE && n_targets == 1) {
            diffs[[target_box]] <- paste(diffs[[target_box]], "+",
                                         decline_term)
          } else {
            fraction_to_target = paste("f", origin_box, "to", target, sep = "_")
            parms <- c(parms, fraction_to_target)
            diffs[[target_box]] <- paste(diffs[[target_box]], "+",
                fraction_to_target, "*", decline_term)
          }
        }
      }
    } #}}}
  } #}}}

  model <- list(diffs = diffs, parms = parms, map = map, spec = spec, use_of_ff = use_of_ff)

  # Create coefficient matrix if possible #{{{
  if (mat) {
    boxes <- names(diffs)
    n <- length(boxes)
    m <- matrix(nrow=n, ncol=n, dimnames=list(boxes, boxes))

    if (use_of_ff == "min") { # {{{ Minimum use of formation fractions
      for (from in boxes) {
        for (to in boxes) {
          if (from == to) { # diagonal elements
            k.candidate = paste("k", from, c(boxes, "sink"), sep = "_")
            k.candidate = sub("free.*bound", "free_bound", k.candidate)
            k.candidate = sub("bound.*free", "bound_free", k.candidate)
            k.effective = intersect(model$parms, k.candidate)
            m[from,to] = ifelse(length(k.effective) > 0,
                paste("-", k.effective, collapse = " "), "0")

          } else {          # off-diagonal elements
            k.candidate = paste("k", from, to, sep = "_")
            if (sub("_free$", "", from) == sub("_bound$", "", to)) {
              k.candidate = paste("k", sub("_free$", "_free_bound", from), sep = "_")
            }
            if (sub("_bound$", "", from) == sub("_free$", "", to)) {
              k.candidate = paste("k", sub("_bound$", "_bound_free", from), sep = "_")
            }
            k.effective = intersect(model$parms, k.candidate)
            m[to, from] = ifelse(length(k.effective) > 0,
                k.effective, "0")
          }
        }
      } # }}}
    } else { # {{{ Use formation fractions where possible
      for (from in boxes) {
        for (to in boxes) {
          if (from == to) { # diagonal elements
            k.candidate = paste("k", from, sep = "_")
            m[from,to] = ifelse(k.candidate %in% model$parms,
                paste("-", k.candidate), "0")
            if(grepl("_free", from)) { # add transfer to bound compartment for SFORB
              m[from,to] = paste(m[from,to], "-", paste("k", from, "bound", sep = "_"))
            }
            if(grepl("_bound", from)) { # add backtransfer to free compartment for SFORB
              m[from,to] = paste("- k", from, "free", sep = "_")
            }
            m[from,to] = m[from,to]
          } else {          # off-diagonal elements
            f.candidate = paste("f", from, "to", to, sep = "_")
            k.candidate = paste("k", from, to, sep = "_")
            k.candidate = sub("free.*bound", "free_bound", k.candidate)
            k.candidate = sub("bound.*free", "bound_free", k.candidate)
            m[to, from] = ifelse(f.candidate %in% model$parms,
              paste(f.candidate, " * k_", from, sep = ""),
              ifelse(k.candidate %in% model$parms, k.candidate, "0"))
            # Special case: singular pathway and no sink
            if (spec[[from]]$sink == FALSE && length(spec[[from]]$to) == 1 && to %in% spec[[from]]$to) {
              m[to, from] = paste("k", from, sep = "_")
            }
          }
        }
      }
    } # }}}
    model$coefmat <- m
  }#}}}

  # Try to create a function compiled from C code if there is more than one observed variable {{{
  # and a compiler is available
  if (length(obs_vars) > 1 & has_compiler()) {

    # Translate the R code for the derivatives to C code
    diffs.C <- paste(diffs, collapse = ";\n")
    diffs.C <- paste0(diffs.C, ";")

    # HS
    diffs.C <- gsub(HS_decline, "(time <= tb ? k1 : k2)", diffs.C, fixed = TRUE)

    for (i in seq_along(diffs)) {
      state_var <- names(diffs)[i]

      # IORE
      if (state_var %in% obs_vars) {
        if (spec[[state_var]]$type == "IORE") {
          diffs.C <- gsub(paste0(state_var, "^N_", state_var),
                          paste0("pow(y[", i - 1, "], N_", state_var, ")"),
                          diffs.C, fixed = TRUE)
        }
      }

      # Replace d_... terms by f[i-1]
      # First line
      pattern <- paste0("^d_", state_var)
      replacement <- paste0("\nf[", i - 1, "]")
      diffs.C <- gsub(pattern, replacement, diffs.C)
      # Other lines
      pattern <- paste0("\\nd_", state_var)
      replacement <- paste0("\nf[", i - 1, "]")
      diffs.C <- gsub(pattern, replacement, diffs.C)

      # Replace names of observed variables by y[i],
      # making the implicit assumption that the observed variables only occur after "* "
      pattern <- paste0("\\* ", state_var)
      replacement <- paste0("* y[", i - 1, "]")
      diffs.C <- gsub(pattern, replacement, diffs.C)
    }

    derivs_sig <- signature(n = "integer", t = "numeric", y = "numeric",
                            f = "numeric", rpar = "numeric", ipar = "integer")

    # Declare the time variable in the body of the function if it is used
    derivs_code <- if (spec[[1]]$type %in% c("FOMC", "DFOP", "HS")) {
      paste0("double time = *t;\n", diffs.C)
    } else {
      diffs.C
    }

    # Define the function initializing the parameters
    npar <- length(parms)
    initpar_code <- paste0(
      "static double parms [", npar, "];\n",
      paste0("#define ", parms, " parms[", 0:(npar - 1), "]\n", collapse = ""),
      "\n",
      "void initpar(void (* odeparms)(int *, double *)) {\n",
      "    int N = ", npar, ";\n",
      "    odeparms(&N, parms);\n",
      "}\n\n")

    # Try to build a shared library
    cf <- try(cfunction(list(func = derivs_sig), derivs_code,
                        otherdefs = initpar_code,
                        verbose = verbose,
                        convention = ".C", language = "C"),
              silent = TRUE)

    if (!inherits(cf, "try-error")) {
      if (!quiet) message("Successfully compiled differential equation model from auto-generated C code.")
      model$cf <- cf
    }
  }
  # }}}

  class(model) <- "mkinmod"
  return(model)
}

#' Print mkinmod objects
#'
#' Print mkinmod objects in a way that the user finds his way to get to its
#' components.
#'
#' @param x An \code{\link{mkinmod}} object.
#' @param \dots Not used.
#' @examples
#'
#'   m_synth_SFO_lin <- mkinmod(parent = list(type = "SFO", to = "M1"),
#'                              M1 = list(type = "SFO", to = "M2"),
#'                              M2 = list(type = "SFO"), use_of_ff = "max")
#'
#'   print(m_synth_SFO_lin)
#'
#' @export
print.mkinmod <- function(x, ...) {
  cat("<mkinmod> model generated with\n")
  cat("Use of formation fractions $use_of_ff:", x$use_of_ff, "\n")
  cat("Specification $spec:\n")
  for (obs in names(x$spec)) {
    cat("$", obs, "\n", sep = "")
    spl <- x$spec[[obs]]
    cat("$type:", spl$type)
    if (!is.null(spl$to) && length(spl$to)) cat("; $to: ", paste(spl$to, collapse = ", "), sep = "")
    cat("; $sink: ", spl$sink, sep = "")
    if (!is.null(spl$full_name)) if (!is.na(spl$full_name)) cat("; $full_name:", spl$full_name)
    cat("\n")
  }
  if (is.matrix(x$coefmat)) cat("Coefficient matrix $coefmat available\n")
  if (!is.null(x$cf)) cat("Compiled model $cf available\n")
  cat("Differential equations:\n")
  nice_diffs <- gsub("^(d.*) =", "\\1/dt =", x[["diffs"]])
  writeLines(strwrap(nice_diffs, exdent = 11))
}
# vim: set foldmethod=marker ts=2 sw=2 expandtab: