Active substances data formats
Active substances as data have to be specified via assessment group (AG) memberships in an AG membership model. For each effect one or more AG membership models can be available, one of which should be chosen in assessments. The AG memberships can be crisp, i.e. a positive list of active substances (with default memberships 1, although it is also allowed to include the negative memberships with membership 0 explicitly) or probabilistic ( \(0 \leq P \leq 1\) ).
Assessment group membership models contain substance membership definitions for a given (health) effect. This data is described using two tables: the assessment group membership models table and the assessment group memberships table. The groups for a specified health effect are defined in the assessment group membership models table. The assessment group memberships table describes the substance memberships (or membership probabilities) in each group.
Download empty dataset template:
Assessment group membership models
This table contains the definitions of the assessment group membership models. Each model contains a id, name, an optional description, and refers to its related health effect.
Accepted table names: AssessmentGroupMembershipModels, AssessmentGroupMembershipModel.
| Name | Type | Description | Aliases | Required |
|---|---|---|---|---|
id |
AlphaNumeric (50) |
The unique identification code of the assessment group membership model. |
id, idModel, Model, idAssessmentGroupModel, AssessmentGroupModel, idGroupMembershipModel, GroupMembershipModel |
Yes |
Name |
AlphaNumeric (100) |
The name of the assessment group membership model. |
Name |
No |
Description |
AlphaNumeric (200) |
Description of the assessment group membership model. |
Description |
No |
idEffect |
AlphaNumeric (50) |
The effect code. |
idEffect, EffectId, Effect |
Yes |
idIndexSubstance |
AlphaNumeric (50) |
The id/code of the index substance. |
idIndexSubstance, idReferenceSubstance, IndexSubstanceId, ReferenceSubstanceId |
No |
Accuracy |
Numeric |
If applicable, the accuracy of the assessment group membership model memberships. |
Accuracy |
No |
Sensitivity |
Numeric |
If applicable, the sensitivity of the assessment group membership model. |
Sensitivity |
No |
Specificity |
Numeric |
If applicable, the specificity of the assessment group membership model. |
Specificity |
No |
Reference |
AlphaNumeric (200) |
External reference(s) to sources containing more information about the assessment group model. |
References |
No |
Assessment group memberships
Substances belong to an assessment group with certainty (probability 1), or the membership are uncertain. This table allows to specify membership probabilities for assessment group membership models. The probability should be a value between zero and one. For example, set to 1 or 0, or prior probabilities, or probabilities or 0/1 values estimated from QSAR, from Molecular Docking or from expert elicitation. The table can contain prior or posterior memberships. Default membership are specified with an empty idSubstance field.
Accepted table names: AssessmentGroupMemberships, AssessmentGroupMembership.
| Name | Type | Description | Aliases | Required |
|---|---|---|---|---|
idGroupMembershipModel |
AlphaNumeric (50) |
The id of the assessment group memberships model or source. |
Model, idModel, idAssessmentGroupMembershipModel, AssessmentGroupMembershipModel, idGroupMembershipModel, GroupMembershipModel, idGroup |
Yes |
idSubstance |
AlphaNumeric (50) |
The code of the substance. |
idSubstance, SubstanceId, SubstanceCode, Substance |
Yes |
GroupMembership |
Numeric |
Probability of the substance for belonging to the assessment group for the effect. If omitted, the default is 1, i.e. certain membership. |
GroupMembership, Membership, MembershipProbability, Probability, AssessmentGroupMembership |
Yes |