Dr. Jill Hagenkord was the keynote speaker at the recent Pathology Informatics Summit in Pittsburgh. She is the chief medical officer of 23andMe. In her lecture, she made reference to the power of the very large genomic database that the company has acquired based on the submission of cell samples by individuals for testing. One of her points referred to the exploration of a genetic basis for cancer of the pancreas. Here are the facts that she presented as shown in her slide relating to the search by a multi-generational pancreatic cancer family with the goal of identifying a possible hereditary cancer syndrome:
- Whole genome scanning (WGS) of an affected individual in the family...[had not been] helpful except for a [identifying a] variant of unknown significance (VUS) in [the] MLH1 DNA repair gene normally association with the Lynch Syndrome, a P603R variant.
- Experts were contacted worldwide for the family with the goal of helping to clarify this VUS in the family but without success.
- 23andMe was then...[contacted] to see if the company database could be of value in elucidating this VUS.
- It was determined that 157 23andMe consumers in the database carry the MLH1, P603R variant.
- It has not yet been determined yet whether this cohort have an increased incidence of pancreatic or colonic cancer.
Variants of unknown significance are DNA changes with too little known information to classify as either pathogenic or benign and it is unknown whether they contribute to a medical condition (see: Ambry Genetics). Here's a quote from a 23andMe publication available online discussing the power of its database to elucidate VUSs of the BRCA1/BRCA2 gene (see: Interpretation of Variants of Unknown Significance in a Large Database of Genotyped and Phenotyped Individuals):
VUS are usually too rare to be amenable to genome-wide association studies and so traditionally have been interpreted with reference to the primary literature (especially for high-penetrance or Mendelian mutations) or by computational methods....While these methods can provide useful insights, they are often limited by a lack of data and presence of false positives in the primary literature and by algorithms that inform about the effect of the variant on the protein and not on the disease state....
Due to the extensive phenotyping of our cohort (over 80 million phenotypic data points) and the large number of rare variants curated on our custom genotyping chip, the method described...[in this article] is applicable to a large number of variants and phenotypes. Our proof-of-principle experiments show that, given sufficient sample size, our database can help uncover the phenotypic effects of high-penetrance variants.