Exposure mixtures settings

Calculation settings

Table 157 Calculation settings for module Exposure mixtures.

Name	Type	Description
Exposure type	ExposureType	The type of exposure considered in the assessment; acute (short term) or chronic (long-term).
Target level	TargetLevelType	Select to express hazard characterisations at external or internal exposure level. For an aggregate assessment, that is dietary and nondietary exposure data are combined, the target dose level is always internal. When only dietary exposures are available, the target dose level is optional, i.c. external or internal.
Exposure calculation method	ExposureCalculationMethod	Method for obtaining exposure estimates. These can be modelled exposures (e.g., external (dietary) exposures or internal exposure estimates obtained by aggregating dietary and non-dietary exposures) or exposure estimates derived from human (bio)monitoring data.
Substance weighting in mixtures	ExposureApproachType	Risk based: exposures in equivalents of the reference substance; standardised: standardised exposures per substance have variance 1; or unweighted exposures: RPFs are equal to 1.
Sparseness parameter	Numeric	Sparseness parameter. Value between 0 (not sparse, many substances) and 1 (sparse, few substances).
SNMU: number of components	Numeric	The number of components to select in SNMU.
Iterations	Numeric	Maximum number of iterations, e.g. 1000.
Seed for pseudo-random number generator.	Numeric	Random seed for initialising matrix W and H.
Convergence criterion	Numeric	Convergence criterion for factorisation algorithm.
Cutoff MCR	Numeric	For selection of individual(day) exposures with maximum cumulative ratio (MCR = total exposure/maximum) above the cutoff.
Cutoff percentage in population ranked on total exposure	Numeric	For selection of individual(day) exposures above the cutoff percentage in the set of individual(day)s ranked on total exposure.
Number of clusters	Numeric	Number of clusters for hierarchical cluster analysis or clustering minimizing within-cluster variance (k-means).
Method	ClusterMethodType	After component selection, a clustering maybe applied on the individuals coefficents matrix V (or H) using K-Means or hierarchical clustering to identify populations sub-groupings.
Determine number of clusters automatically	Boolean	For hierarchical clustering the number of clusters is determined automatically.
Network analysis type	NetworkAnalysisType	Network analysis on the substance x component (U) matrix.
Apply log transformation	Boolean	Network analysis: apply log transform on exposures.