set_fixed_css: Prepare steady-state plasma concentrations for sensitivity...

In GeoTox: Spatiotemporal Mixture Risk Assessment

Prepare steady-state plasma concentrations for sensitivity analysis

Description

Create the 'fixed_css' table in the GeoTox database, which contains values of steady-state plasma concentrations (C_{ss}) for sensitivity analysis.

Usage

set_fixed_css(GT, substance_order = NULL)

Arguments

GT	GeoTox object.
substance_order	Named list specifying order of substance evaluation (default NULL).

Details

Several tables are required in the GeoTox database before the 'fixed_css' table can be created. Typically, set_fixed_css() is called after using sample_simulated_css(), at which point all required tables will be present.

The resulting 'fixed_css' table is used for sensitivity analysis where one parameter is allowed to vary at a time while all other parameters are held constant at fixed values.

The substance_order argument can be used to specify the order in which substances are evaluated when sampling C_{ss} values. The default is to evaluate substances in the order they appear in the 'substance' table. However, if a different order is needed for some reason (e.g., replication of results from a previous GeoTox implementation), the user can provide a named list where the name is the column in the 'substance' table to use for ordering (e.g., "casn") and the value is a vector of substance identifiers in the desired order. If provided, the substance_order will be added to the GeoTox object parameter list, GT$par. This is only applicable to the "params" section described below.

Table 'fixed_css' columns:

age

Median pre-simulated C_{ss} values are computed by age group for each substance, then the median C_{ss} values are assigned to each individual based on their age group.

weight

Median pre-simulated C_{ss} values are computed by weight category for each substance, then the median C_{ss} values are assigned to each individual based on their weight category.

params

Individuals are assigned the median age of their location and pre-simulated C_{ss} values for the "Normal" weight category are sampled.

other

Median sampled C_{ss} values are computed across all substances for each location after mean-imputation of missing C_{ss} values for each substance and location. The median C_{ss} values are assigned to each individual based on their location.

Value

The updated GeoTox object, invisibly.

See Also

sample_simulated_css(), simulate_population()

Examples

# Create required tables
sample_df <- tibble::tribble(
  ~FIPS, ~age, ~weight,
  10000, 25, "Normal",
  10000, 35,  "Obese",
  20000, 50, "Normal"
)
exposure_df <- tibble::tribble(
  ~FIPS, ~casn, ~route, ~mean, ~sd,
  10000, "00-00-1", "inhalation", 10, 1,
  10000, "00-00-2", "inhalation", 20, 1,
  20000, "00-00-1", "inhalation", 30, 1,
  20000, "00-00-2", "inhalation", 40, 1
)
# Note: normally the css_df would have many more rows for each combination of
# the non-'css' columns to allow for sampling.
css_df <- tibble::tribble(
  ~casn, ~age_lb, ~age_ub, ~weight, ~css,
  "00-00-1",  0, 49, "Normal",  1,
  "00-00-1", 50, 99, "Normal",  2,
  "00-00-1",  0, 49,  "Obese", 11,
  "00-00-1", 50, 99,  "Obese", 12,
  "00-00-2",  0, 49, "Normal", 21,
  "00-00-2", 50, 99, "Normal", 22,
  "00-00-2",  0, 49,  "Obese", 31,
  "00-00-2", 50, 99,  "Obese", 32
)
GT <- GeoTox() |>
  set_sample(sample_df) |>
  add_exposure(exposure_df) |>
  simulate_exposure() |>
  set_simulated_css(css_df) |>
  sample_simulated_css()

# Set fixed C_ss values
GT <- GT |> set_fixed_css()

# Open a connection to GeoTox database
con <- get_con(GT)

# Look at created tables
# Note: the 'age', 'weight', 'params', and 'other' columns of the
# 'fixed_css' table contain the C_ss values for sensitivity analysis.
# For example, the 'age' column doesn't contain ages, but C_ss values.

dplyr::tbl(con, "fixed_css") |> dplyr::collect()

dplyr::tbl(con, "concentration") |> dplyr::collect()

dplyr::tbl(con, "sample") |> dplyr::collect()

dplyr::tbl(con, "location") |> dplyr::collect()

dplyr::tbl(con, "substance") |> dplyr::collect()

# Clean up example
DBI::dbDisconnect(con)
file.remove(GT$db_info$dbdir)

GeoTox documentation built on May 20, 2026, 1:07 a.m.