Building a CCD chemical profile

This article builds one Chemical Dashboard profile for benzene. It uses DTXSID7020182 as the join key. Keep units, sources, and evidence categories when you combine results.

Start with chemical identity

library(ComptoxR)
library(dplyr)

dtxsid <- "DTXSID7020182"
detail <- ct_chemical_detail_search(dtxsid)

glimpse(detail)
detail |>
  select(dtxsid, preferredName, casrn, molFormula, averageMass, smiles)

A verified production response has one row and includes dtxsid, preferredName, casrn, molFormula, averageMass, and smiles, with additional identifiers and Dashboard summary fields.

Add physical-chemical properties

Property names are service values. Inspect them before you build a large query.

ct_chemical_property_experimental_name()
ct_chemical_property_predicted_name()

water_summary <- ct_chemical_property_summary_search(
  dtxsid,
  propName = "Water Solubility"
)
water_experimental <- ct_chemical_property_summary_experimental_search(
  dtxsid,
  propName = "Water Solubility"
)
water_predicted <- ct_chemical_property_summary_predicted_search(
  dtxsid,
  propName = "Water Solubility"
)

water_summary |>
  select(propName, unit, experimentalMedian, experimentalCount,
         predictedMedian, predictedCount)
water_experimental |> select(propValue, sourceName, publicSourceUrl)
water_predicted |> select(propValue, sourceName, qmrfUrl)

The summary combines counts and simple statistics. Experimental rows give reported values and sources. Predicted rows give model values and, when available, model-report links. Do not compare values until you confirm their units and basis.

Add environmental fate

The fate wrappers use the same propName pattern. Some chemicals have only experimental or only predicted evidence.

bcf_summary <- ct_chemical_fate_summary_search(
  dtxsid,
  propName = "Bioconcentration Factor"
)
bcf_experimental <- ct_chemical_fate_experimental_search(
  dtxsid,
  propName = "Bioconcentration Factor"
)
bcf_predicted <- ct_chemical_fate_predicted_search(
  dtxsid,
  propName = "Bioconcentration Factor"
)

bcf_summary |>
  select(propName, unit, experimentalMedian, experimentalCount,
         predictedMedian, predictedCount)
bcf_experimental |> select(propValue, sourceName, publicSourceUrl)
bcf_predicted |> select(propValue, sourceName, qmrfUrl)

Other current fate property names include "Biodeg. Half-Life", "Atmos. Hydroxylation Rate", and "Soil Adsorp. Coeff. (Koc)".

Add hazard evidence

The CCD ToxVal endpoint is a fast public API path. Select a small set of fields and keep the evidence metadata.

toxval <- ct_hazard_toxval_search(dtxsid)

toxval |>
  filter(toxvalTypeSuperCategory %in% c("Point of Departure", "Reference Value")) |>
  select(dtxsid, source, toxvalType, toxvalTypeSuperCategory,
         toxvalNumeric, toxvalUnits, speciesCommon, studyType,
         quality, qcCategory)

genetox <- ct_hazard_genetox_search(dtxsid)
cancer <- ct_hazard_cancer_search(dtxsid)
toxref <- ct_hazard_toxref_data_search(dtxsid)

toxvalNumeric has meaning only with toxvalUnits. toxvalType states the value type, while toxvalTypeSuperCategory groups evidence such as points of departure and reference values. Source, species, study type, quality, and QC fields help a reviewer judge the evidence.

Retrieve IVIVE parameters

ivive <- ct_hazard_adme_ivive_search(dtxsid)
ivive |>
  select(dtxsid, description, measured, predicted, unit, model,
         species, reference)

The current response separates measured and predicted values. Keep unit, model, species, and reference with each parameter.

Add exposure estimates

SEEM general exposure is the primary compact exposure result.

seem <- ct_exposure_seem_general_search(dtxsid)
seem |>
  select(dtxsid, productionVolume, units, starts_with("probability"))

production <- ct_exposure_ccd_production_volume_search(dtxsid)
production |> select(dtxsid, name, amount)

The SEEM probabilities describe exposure-use categories. They are screening estimates, not measured individual exposure. Production-volume records are a short supporting line of evidence.

Start compact, then inspect assay records

bioactivity_summary <-
  ct_bioactivity_data_summary_search_by_dtxsid(dtxsid)
bioactivity_summary |>
  select(dtxsid, activeMc, totalMc, activeSc, totalSc, ntested, nhit,
         cytotoxMedianUm)

bioactivity_detail <- ct_bioactivity_data_search_by_dtxsid(dtxsid)
glimpse(bioactivity_detail)

The summary gives assay counts and cytotoxicity context before you work with detailed assay records.

Know which calls are bulk

Use bulk wrappers for chemical detail, experimental and predicted properties, fate records, CCD ToxVal, cancer, genetox, SEEM general exposure, and detailed bioactivity data. The property and fate summary wrappers, IVIVE, production volume, and bioactivity summary wrappers require one DTXSID at a time. Use a small, rate-conscious iteration only when no bulk wrapper exists. The next article shows both patterns.