Table 1:

Simulation Parameters

Parameter	Interpretation	Possible Values
$q$	Number of sets of biomarkers	15, 20, 50
$q_{OD},q_{TD}$	Number of one- or two-dimensional sets	0, 4, 6
$m_s$	Number of biomarkers per set	15, 20, 50
$m_M$	Number of mediating biomarkers in the mediating set	1, 3, 5
$m_E$	Number of noise biomarkers in the mediating set associated only with exposure	0, 2, 4
$m_Y$	Number of noise biomarkers in the mediating set associated only with outcome	0, 2, 4
$m_{EY}$	Number of noise biomarkers in the mediating set with half associated only with the exposure and half associated only with the outcome	0, 2, 4
$m_D$	Number of associated biomarkers in one- or two-dimensional sets	6
${\rho }_M$	Correlation between biomarkers within a set	0, 0.25, 0.5
${\beta }_j$	Non-null effect of exposure on metabolite	0.065A, 0.045B
${\gamma }_j^{*}$	Non-null effect of metabolite on outcome	0.065 A, 0.085B

^AModels with a continuous outcome

^BModels with a binary outcome

Bold values indicate default settings.