About HEX


As the pace of ascertaining genetic data accelerates and the volume of data grows, researchers face the challenge of understanding the impact of observed genetic variability. Is a given sequence variant neutral, protective, or pathogenic? Together with other resources, HEX aims to enable scientists to more confidently interpret the biological consequences of a given variant.

Genetic studies fall into two general categories. Association studies involve the sequencing of large cohorts of cases and controls; family studies involve sequencing of multiple members of a family in whom a disease segregates. The former requires as many participants as possible to achieve statistical power to detect associations between variants and disease/phenotype; the latter requires sorting through the variants that differ between affected and unaffected family members and determining which are likely to be pathogenic. HEX is a resource designed to aid in the sorting, or filtering, primarily in family based studies. However, in certain scenarios HEX may be useful for association studies, as well (see below).

A common approach to filtering variants is to use publicly available databases, namely ExAC and gnomAD, in order to check for presence or absence of the variant at hand in a larger population. These resources contain copious amounts of human genetic data; however, this data comes with scant phenotyping of the people who donated the DNA. In particular, information such as age and health status is critical for genetic studies of age-related neurodegeneration.

The HEX database provides a unique reference cohort in that it includes genetic data only from people who lived to age 60 or older, and who had no neurodegenerative disease diagnosis, or overt neurodegenerative disease pathology upon postmortem examination. Therefore, the HEX cohort is extremely unlikely to contain variants associated with early onset neurodegenerative disease, which typically manifests before age 60. In other words, a novel exonic sequence variant found in an association or family study is unlikely to play a major role in early onset neurodegenerative disease if it also occurs in HEX.

That said, some variants found in HEX may still be involved in neurodegeneration. Many cases of Alzheimer’s and related disease are polygenic, whereby multiple variants together determine a person’s individual risk and age at onset. While not causing early onset neurodegeneration as an autosomal-dominant mutation would, some variants present in HEX may contribute to disease risk. Some HEX variants may be associated with incomplete penetrance, or with a later onset. For example, the R47H AD risk-associated variant in TREM2 does appear in HEX.

Although HEX was designed primarily to aid in the interpretation of family based studies, there are scenarios in which it may add useful information to interpreting association studies. For example, in a GWAS-type study a variant is found to be associated with decreased risk of AD and with decreased expression of a certain gene. If HEX contains multiple loss-of-function variants in that gene, this information would support the interpretation that the variant is protective.

At present, HEX contains genetic data from a small number of pathologically confirmed elderly, but it will become a more powerful tool as sample size grows.

For guidelines on interpreting genetic variability see MacArthur et al., 2014; for in-depth discussion on analysing large datasets for pathogenic variability, see Walsh et al., 2017, and Minikel et al., 2016


The generation of the HEX data was supported in part by the Intramural Research Programs of the National Institute on Aging and the National Institute of Neurological Disease and Stroke, National Institutes of Health (Department of Health and Human Services Project Numbers Z01 AG000950-10 and Z01 AG000949-02), Alzheimer’s Research UK (ARUK), and an anonymous charitable foundation.  Dr. Guerreiro and Dr. Bras are supported by fellowships from the Alzheimer’s Society. The generation of the HEX data was also supported in part by the Wellcome Trust/MRC Joint Call in Neurodegeneration award (WT089698) to the UK Parkinson’s Disease Consortium, whose members are from the UCL Institute of Neurology, the University of Sheffield, and the MRF Protein Phosphorylation Unit at the University of Dundee.