Registration Dossier

Administrative data

Endpoint:
acute toxicity: oral
Type of information:
(Q)SAR
Adequacy of study:
key study
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Data source

Reference
Title:
T.E.S.T. Oral rat LD50 (version 4.2.1)
Year:
2016
Bibliographic source:
Martin, T; Toxicity Estimation Software Tool (TEST); U.S. Environmental Protection Agency, Washington, DC, (2016)

Materials and methods

Test guideline
Guideline:
other: REACH Guidance on QSARs R.6
Principles of method if other than guideline:
Model developer(s) and contact details:
U.S. Environmental Protection Agency

Reference(s) to main scientific papers and/or software package
Martin, T; Toxicity Estimation Software Tool (TEST); U.S. Environmental Protection Agency, Washington, DC, (2016)
https://www.epa.gov/chemical-research/toxicity-estimation-software-tool-test

Availability of information about the model:
Freely available through the EPA web site:
https://www.epa.gov/chemical-research/toxicity-estimation-software-tool-test

Endpoint:
Endpoint represents the amount of the chemical (mass of the chemical per body weight of the rat) which when orally ingested kills half of rats

Test material

Reference
Name:
Unnamed
Type:
Constituent
Specific details on test material used for the study:
C(=C\F)\C(F)(F)F

Results and discussion

Effect levels
Sex:
not specified
Dose descriptor:
LD50
Effect level:
ca. 1 240 mg/kg bw

Applicant's summary and conclusion

Interpretation of results:
Category 4 based on GHS criteria
Conclusions:
A LD50 (-log(LD50 mol/kg) of 2.0 (i.e. about 1240 mg/kg) can be considered a good precautionary estimation of oral rat toxicity for the target in the T.E.S.T. Oral rat LD50 v 4.2.1 model.
Executive summary:

Predicted LD50 as mg/kg is 1272 mg/kg. The results of the model can’t be considered reliable in terms of numerical value because it is likely that the model overestimate the toxicity. Despite the low reliability of the prediction, this estimation can be used in a conservative way. Closest neighbors in the training set (151-67-7 and 6186-91-0) over-predicted (with an error in prediction higher than 1 log unit) and the MAE of similar compounds in the training set higher than the error observed for the whole dataset. Indeed, because highly similar compounds identified in the dataset are over-predicted, the model is also likely to predict the target as more toxic than its real value. A LD50 (-log(LD50 mol/kg) of 2.0 (i.e. about 1240 mg/kg) can be considered a good precautionary estimation of oral rat toxicity for the target.