Table 1.

Glossary.

Genetics
Heritability	The proportion of variance of a phenotype that is attributable to genetic factors.
Genetic correlation	The degree to which two phenotypes are influenced by the same genetic variation.
GWAS	Genome-wide association study: mass-univariate analysis to relate common variation over the entire length of the DNA to a phenotype of interest.
SNP	Single nucleotide polymorphism: a (common) genetic variation in the DNA sequence where different alleles (nucleotides) can exist in the population.
Polygenic	Influenced by many genetic variants (i.e., hundreds, or thousands of genes), as opposed to monogenic (influenced by a single gene, or single genetic variant).
Mendelian Randomisation	Hypothesis-driven method aimed at inferring causality from (cross-sectional) associations between a genetic variant and two or more phenotypes. E.g. to test whether a modifiable behavioural or neural trait potentially mediates the effect of a genetic variant on a disease [95].
LD-score regression	Linkage-disequilibrium score regression: method to calculate genetic correlations on the basis of GWAS output (i.e., “summary statistics”), given the relationship of the statistics to each variant’s linkage disequilibrium pattern [25]
Neuroimaging
MRI	Magnetic Resonance Imaging
Functional MRI	MRI acquisition method to estimate regional brain activation based on local blood-oxygen level dependent (BOLD) signal.
Diffusion MRI	MRI acquisition method to measure microstructural tissue properties based on direction and amount of diffusion of water molecules. Most often used for investigating white matter fibres.
Functional connectivity	The degree to which two or more brain regions show similar activation patterns over time, based on the correlation or mutual dependence of their BOLD time-series.
Multivariate methods
PCA	Principal Component Analysis: data-driven data reduction method to extract maximally uncorrelated components (i.e., “factors”) from many variables.
ICA	Independent Component Analysis: data-driven data reduction method and source identification method, which extracts maximally independent components (i.e., “factors”) from many variables.
CCA	Canonical Correlation Analysis: method to extract modes (here:“factors”) across two or more sets of variables (e.g., MRI and behavioural variables), such that the variables within a mode are maximally correlated.
SEM	Structural Equation Modelling: data reduction method to fit a priori factor structures to data and extract these factors. Can be confirmatory (1 model is tested) or exploratory (multiple a priori models are tested and compared).

^*Note: For consistency and clarity, throughout the paper the term “factors” is used to describe all kinds of factors, components, dimensions, sources, or modes, even if the term “factors” is unusual for the particular method that was used. For the purpose of the present paper, the interpretation is the same across these terms.