General population genetics-neuroimaging cohorts gather two types of data: brain imaging and genetics. These unique tools make it possible to research correlations between genetic variants and imaged brain features, and more generally, to study the influence of genetics and the environment on the variance of brain features observed in normal and pathological populations. By analyzing such cohorts, the researchers hope to discover genetic associations with earlier or more reliable image characteristics, as compared to the diagnostics observable at the individual clinical level.

The UK Biobank cohort, comprising 16,304 subjects at the time of the study, is currently the best example of this ambitious approach. For each subject, the researchers measured:

• 1,051,316 haplotypes, i.e. sequences of several contiguous variants inherited from the same parent and generally transmitted together;
• and the opening (measured in millimeters) of 123 cerebral sulci (linear depressions on the surface of the cortex).

They then looked for correlations between these two sets of measurements, equal to 1,051,316 × 123 combinations! They could thus identify several genomic regions of interest (on chromosomes 9, 12, and 16) associated with sulci openings. This result was replicated in approximately 5,000 non-cohort subjects.

The researchers could thus map sulci across the whole brain that show variability in openings related at least partially to genomic regions. They were also able to compare their haplotypic approach to classical association methods using SNPs only where each genetic variant is treated individually. In particular, they could demonstrate that their genome-wide haplotype analyses are more sensitive and represent an interesting, statistically sound alternative. 

Learn more about the UK Biobank cohort.