httk (version 2.1.0)

get_cheminfo: Retrieve chemical information from HTTK package

Description

This function provides the information specified in "info=" (can be single entry or vector) for all chemicals for which a toxicokinetic model can be parameterized for a given species. Since different models have different requirements and not all chemicals have complete data, this function will return different number of chemicals depending on the model specififed.

Usage

get_cheminfo(
  info = "CAS",
  species = "Human",
  fup.lod.default = 0.005,
  model = "3compartmentss",
  default.to.human = FALSE,
  median.only = FALSE,
  fup.ci.cutoff = TRUE,
  clint.pvalue.threshold = 0.05,
  suppress.messages = FALSE
)

Value

vector/data.table

Table (if info has multiple entries) or vector containing a column for each valid entry specified in the argument "info" and a row for each chemical with sufficient data for the model specified by argument "model":

Column	Description	units
Compound	The preferred name of the chemical compound	none
CAS	The preferred Chemical Abstracts Service Registry Number	none
DTXSID	DSSTox Structure ID (http://comptox.epa.gov/dashboard)	none
logP	The log10 octanol:water partition coefficient	log10 unitless ratio
MW	The chemical compound molecular weight	g/mol
pKa_Accept	The hydrogen acceptor equilibria concentrations	logarithm
pKa_Donor	The hydrogen donor equilibria concentrations	logarithm
[SPECIES].Clint	(Primary hepatocyte suspension) intrinsic hepatic clearance	uL/min/10^6 hepatocytes
[SPECIES].Clint.pValue	Probability that there is no clearance observed.	none
[SPECIES].Funbound.plasma	Chemical fraction unbound in presence of plasma proteins	unitless fraction
[SPECIES].Rblood2plasma	Chemical concentration blood to plasma ratio	unitless ratio

Arguments

info

A single character vector (or collection of character vectors) from "Compound", "CAS", "DTXSID, "logP", "pKa_Donor"," pKa_Accept", "MW", "Clint", "Clint.pValue", "Funbound.plasma","Structure_Formula", or "Substance_Type". info="all" gives all information for the model and species.

species

Species desired (either "Rat", "Rabbit", "Dog", "Mouse", or default "Human").

fup.lod.default

Default value used for fraction of unbound plasma for chemicals where measured value was below the limit of detection. Default value is 0.0005.

model

Model used in calculation, 'pbtk' for the multiple compartment model, '1compartment' for the one compartment model, '3compartment' for three compartment model, '3compartmentss' for the three compartment model without partition coefficients, or 'schmitt' for chemicals with logP and fraction unbound (used in predict_partitioning_schmitt).

default.to.human

Substitutes missing values with human values if true.

median.only

Use median values only for fup and clint. Default is FALSE.

fup.ci.cutoff

Cutoff for the level of uncertainty in fup estimates. This value should be between (0,1). Default is `NULL` specifying no filtering.

clint.pvalue.threshold

Hepatic clearance for chemicals where the in vitro clearance assay result has a p-values greater than the threshold are set to zero.

suppress.messages

Whether or not the output messages are suppressed.

Author

John Wambaugh, Robert Pearce, and Sarah E. Davidson

Details

When default.to.human is set to TRUE, and the species-specific data, Funbound.plasma and Clint, are missing from chem.physical_and_invitro.data, human values are given instead.

In some cases the rapid equilbrium dailysis method (Waters et al., 2008) fails to yield detectable concentrations for the free fraction of chemical. In those cases we assume the compound is highly bound (that is, Fup approaches zero). For some calculations (for example, steady-state plasma concentration) there is precendent (Rotroff et al., 2010) for using half the average limit of detection, that is, 0.005 (this value is configurable via the argument fup.lod.default). We do not recomend using other models where quantities like partition coefficients must be predicted using Fup. We also do not recomend including the value 0.005 in training sets for Fup predictive models.

Note that in some cases the Funbound.plasma and the intrinsic clearance are provided as a series of numbers separated by commas. These values are the result of Bayesian analysis and characterize a distribution: the first value is the median of the distribution, while the second and third values are the lower and upper 95th percentile (that is qunatile 2.5 and 97.5) respectively. For intrinsic clearance a fourth value indicating a p-value for a decrease is provided. Typically 4000 samples were used for the Bayesian analusis, such that a p-value of "0" is equivale to "<0.00025". See Wambaugh et al. (2019) for more details. If argument meadian.only == TRUE then only the median is reported for parameters with Bayesian analysis distributions. If the 95 credible interval is larger than fup.ci.cutoff (defaults to NULL) then the Fup is treated as too uncertain and the value NA is given.

References

Rotroff, Daniel M., et al. "Incorporating human dosimetry and exposure into high-throughput in vitro toxicity screening." Toxicological Sciences 117.2 (2010): 348-358.

Waters, Nigel J., et al. "Validation of a rapid equilibrium dialysis approach for the measurement of plasma protein binding." Journal of pharmaceutical sciences 97.10 (2008): 4586-4595.

Wambaugh, John F., et al. "Assessing toxicokinetic uncertainty and variability in risk prioritization." Toxicological Sciences 172.2 (2019): 235-251.

Examples

# \donttest{
# List all CAS numbers for which the 3compartmentss model can be run in humans: 
get_cheminfo()

get_cheminfo(info=c('compound','funbound.plasma','logP'),model='pbtk') 
# See all the data for humans:
get_cheminfo(info="all")

TPO.cas <- c("741-58-2", "333-41-5", "51707-55-2", "30560-19-1", "5598-13-0", 
"35575-96-3", "142459-58-3", "1634-78-2", "161326-34-7", "133-07-3", "533-74-4", 
"101-05-3", "330-54-1", "6153-64-6", "15299-99-7", "87-90-1", "42509-80-8", 
"10265-92-6", "122-14-5", "12427-38-2", "83-79-4", "55-38-9", "2310-17-0", 
"5234-68-4", "330-55-2", "3337-71-1", "6923-22-4", "23564-05-8", "101-02-0", 
"140-56-7", "120-71-8", "120-12-7", "123-31-9", "91-53-2", "131807-57-3", 
"68157-60-8", "5598-15-2", "115-32-2", "298-00-0", "60-51-5", "23031-36-9", 
"137-26-8", "96-45-7", "16672-87-0", "709-98-8", "149877-41-8", "145701-21-9", 
"7786-34-7", "54593-83-8", "23422-53-9", "56-38-2", "41198-08-7", "50-65-7", 
"28434-00-6", "56-72-4", "62-73-7", "6317-18-6", "96182-53-5", "87-86-5", 
"101-54-2", "121-69-7", "532-27-4", "91-59-8", "105-67-9", "90-04-0", 
"134-20-3", "599-64-4", "148-24-3", "2416-94-6", "121-79-9", "527-60-6", 
"99-97-8", "131-55-5", "105-87-3", "136-77-6", "1401-55-4", "1948-33-0", 
"121-00-6", "92-84-2", "140-66-9", "99-71-8", "150-13-0", "80-46-6", "120-95-6",
"128-39-2", "2687-25-4", "732-11-6", "5392-40-5", "80-05-7", "135158-54-2", 
"29232-93-7", "6734-80-1", "98-54-4", "97-53-0", "96-76-4", "118-71-8", 
"2451-62-9", "150-68-5", "732-26-3", "99-59-2", "59-30-3", "3811-73-2", 
"101-61-1", "4180-23-8", "101-80-4", "86-50-0", "2687-96-9", "108-46-3", 
"95-54-5", "101-77-9", "95-80-7", "420-04-2", "60-54-8", "375-95-1", "120-80-9",
"149-30-4", "135-19-3", "88-58-4", "84-16-2", "6381-77-7", "1478-61-1", 
"96-70-8", "128-04-1", "25956-17-6", "92-52-4", "1987-50-4", "563-12-2", 
"298-02-2", "79902-63-9", "27955-94-8")
httk.TPO.rat.table <- subset(get_cheminfo(info="all",species="rat"),
 CAS %in% TPO.cas)
 
httk.TPO.human.table <- subset(get_cheminfo(info="all",species="human"),
 CAS %in% TPO.cas)
# }