Additions and corrections

Assessing confidence in predictions made by knowledge-based systems

Philip N. Judson, Susanne A. Stalford and Jonathan Vessey

Toxicol. Res., 2013, 2 (DOI: 10.1039/C2TX20037F). Amendment published 16th May 2014.

The authors thank Alison Callahan and Mukesh Patel for drawing attention to the following errors in their manuscript:

1. The values under “Probable”, “Plausible”, and “Improbable” for Ames mutagenicity dataset 1 in Table 10 should be 0.15, 0.13, and 0.17, respectively.

2. The values under column 1 (for Ames mutagenicity) in Table 11 should be: Probable, 15.5; Plausible, 512; Equivocal, 1.5; Doubted, - ; Improbable, 1.17; Number of predictions, 4074; Aggregate deviation, 0.13; Veracity, 0.87 (this last value is correct in the published table).

3. The definition of Rc below the Veracity equation at the bottom of column 1 on page 75 should be Rc = (C-c)/(C-1) [i.e. not c/(C-1)].

4. In the list of “Let” statements in column 2 on page 76: “... X be the number of compounds predicted negative that were actually positive” should be “... z be the number of compounds predicted negative that were actually positive”; “... Z be the number of positive predictions” should be “... X be the number of positive predictions”.

5. The statement towards the bottom of the second column on page 76, “Since Z + X = N:” should be “Since X + Y = N:”.

The Royal Society of Chemistry apologises for these errors and any consequent inconvenience to authors and readers.

Back to article