identifier |
required |
unique identification of the body of evidence |
string |
arbitrary expression |
source |
required |
type of the body of evidence |
string |
one of the following: 'expert','qsar','in vitro', 'positive alert', 'negative alert' |
result |
required |
result generated by the body of evidence |
number |
a) a single number always in respect to the positive class (1 means positive) |
|
|
|
number |
b) a single probability score in respect to the positive class (e.g.: 0.75 probability of being positive) |
|
|
|
list |
c) list with two probability scores collected for the negative and positive results (starting with negative) |
reliability |
required |
reliability score of the body of evidence |
number |
a) a single number representing the accuracy score, on the scale from 0 to 100 or from 0 to 1 |
|
|
|
dictionary |
b) a dictionary with two values, on the scale from 0 to 1, representing the negative and the positive reliability scores (starting with negative) |
|
|
|
list |
c) a list with two values, on the scale from 0 to 1, representing the negative and the positive reliability scores (starting with negative) |
relevance |
optional |
expert-defined relevance term linking the evidence source (not the result itself) to the assessment question; the default setting is “certain” meaning that the evidence source is completely relevant to the endpoint |
string |
a) 'certain' equivalent to 1 |
|
|
|
string |
b) 'probable' equivalent to 0.9 |
|
|
|
string |
c) 'plausible' equivalent to 0.75 |
|
|
|
string |
d) 'equivocal' equivalent to 0.5 |
|
|
|
string |
e) 'doubted' equivalent to 0.25 |
|
|
|
string |
f) 'improbable' equivalent to 0.1 |
|
|
|
string |
g) 'impossible' equivalent to 0 |
weight |
optional |
expert-defined weight of the evidence body, as compared to other evidence bodies considered for combination; the default setting is 1 |
integer number |
arbitrary integer number (e.g.: 1, 2, 3, …) |