httk (version 2.6.0)

load_dawson2021: Load CLint and Fup QSPR predictions from Dawson et al. 2021.

Description

This function returns an updated version of chem.physical_and_invitro.data that includes Clint and Fup predictions from the Random Forest quantitative structure-property relationship (QSPR) models developed and presented in Dawson et al. 2021, included in table dawson2021.

Usage

load_dawson2021(
  overwrite = FALSE,
  exclude_oad = TRUE,
  chem_include = NULL,
  target.env = .GlobalEnv
)

Value

data.frame

An updated version of chem.physical_and_invitro.data.

Arguments

overwrite

Only matters if load.image=FALSE. If overwrite=TRUE then existing data in chem.physical_and_invitro.data will be replaced by any predictions in Dawson et al. (2021) that is for the same chemical and property. If overwrite=FALSE (DEFAULT) then new data for the same chemical and property are ignored. Funbound.plasma values of 0 (below limit of detection) are overwritten either way.

exclude_oad

Include the chemicals only within the applicability domain. If exclude_oad=TRUE (DEFAULT) chemicals outside the applicability domain do not have their predicted values loaded.

chem_include

A vector of CAS numbers indicating only the chemicals to be included in the loading process. If set to `NULL` all applicable chemicals are loaded. (Default is `NULL`.)

target.env

The environment where the new chem.physical_and_invitro.data is loaded. Defaults to global environment.

Author

Sarah E. Davidson

Details

Because Clint and Fup are the only measurements required for many HTTK models, changing the number of chemicals for which a value is available will change the number of chemicals which are listed with the get_cheminfo command. Use the command reset_httk to return to the initial (measured only) chem.physical_and_invitro.data (for all parameters).

References

dawson2021qsarhttk

See Also

reset_httk

get_cheminfo

Examples

# \donttest{
# Count how many chemicals for which HTTK is available without the QSPR:
num.chems <- length(get_cheminfo())
print(num.chems)

# For chemicals with Dawson et al. (2021) Clint and Fup QSPR predictions, 
# add them to our chemical information wherever measured values are 
# unavailable:
load_dawson2021()
# For chemicals with Dawson et al. (2021) QSPR predictions, add them to
# our chemical information -- overwriting measured values where we had them:
load_dawson2021(overwrite=TRUE)

# Let's see how many chemicals we have now with the Dawson et al. (2021) 
# predictions loaded:
length(get_cheminfo()) 

# Now let us reset the chemical data to the initial version:
reset_httk()

# We should be back to our original number:
num.chems == length(get_cheminfo())

# Demonstrate loading data for specific chemicals:
#
# Find chemicals with a clint and no fup:
subset(chem.physical_and_invitro.data,!is.na(Human.Clint) & Human.Funbound.plasma==0)$CAS
chem1 <- "32598-13-3"
chem2 <- "2971-36-0"
# Take a look at what parameterize_steadystate gives (working from a default fup of 0.005):
a1 <- parameterize_steadystate(chem.cas=chem1)
a2 <- parameterize_steadystate(chem.cas=chem2)

# load Dawson for this chemical:
load_dawson2021(chem_include=chem1)
# Check values, only fup for the first chemical should change:
a3 <- parameterize_steadystate(chem.cas=chem1)
a4 <- parameterize_steadystate(chem.cas=chem2)
# All tests should be true:
a1[["Clint"]] == a3[["Clint"]]
a1[["Funbound.plasma"]] != a3[["Funbound.plasma"]]
a2[["Clint"]] == a4[["Clint"]]
a2[["Funbound.plasma"]] == a4[["Funbound.plasma"]]

# load Dawson for this chemical, but allow it to overwrite the clint:
load_dawson2021(chem_include=chem1, overwrite=TRUE)
# Check values, both clint and fup for the first chemical should change:
a5 <- parameterize_steadystate(chem.cas=chem1)
a6 <- parameterize_steadystate(chem.cas=chem2)
# All tests should be true:
a1[["Clint"]] != a5[["Clint"]]
a1[["Funbound.plasma"]] != a5[["Funbound.plasma"]]
a2[["Clint"]] == a6[["Clint"]]
a2[["Funbound.plasma"]] == a6[["Funbound.plasma"]]

# Load Dawson for all chemicals, fup should change for second chemical:
load_dawson2021()
a7 <- parameterize_steadystate(chem.cas=chem1)
a8 <- parameterize_steadystate(chem.cas=chem2)
# All tests should be true:
a1[["Clint"]] != a7[["Clint"]]
a1[["Funbound.plasma"]] != a7[["Funbound.plasma"]]
a2[["Clint"]] == a8[["Clint"]]
a2[["Funbound.plasma"]] != a8[["Funbound.plasma"]]

# load Dawson for this chemical, but allow it to overwrite all clints:
load_dawson2021(overwrite=TRUE)
# Both clint and fup should now be changed for second chemical:
a9 <- parameterize_steadystate(chem.cas=chem1)
a10 <- parameterize_steadystate(chem.cas=chem2)
# All tests should be true:
a1[["Clint"]] != a9[["Clint"]]
a1[["Funbound.plasma"]] != a9[["Funbound.plasma"]]
a2[["Clint"]] != a10[["Clint"]]
a2[["Funbound.plasma"]] != a10[["Funbound.plasma"]]
# }