Adding this a separate answer because it grew long (surprise!), but it started out as a reply to Greg. So...hi, Greg!
Personally, I'd do some research before making Helix my first choice for anything genealogical. There's really nothing about their products and services that are geared toward genealogy and recent ancestry. National Geographic's Genographic 2.0 uses them as a test facility (having switched after many years with Family Tree DNA) as does Spencer Wells's Insitome...but in neither of those cases does anything more recent than the anthropological timeframe come into play. After Helix testing, for example, and for only a few dollars, Insitome will return information about your percentage of Neanderthal DNA.
I never responded to Levi's May 11 post (sorry, Levi), so probably should take a moment for an overview. To avoid a rabbit-hole dive, just stop now with the "do some research" advice.
Helix, technically Helix Opco LLC, started operations in the summer of 2017. A large investor of the start-up was, and still is, Illumina...one of, really, only two major players in the DNA technology, equipment, and chemistry industry. For both Illumina and Thermo Fisher Scientific, the revenue comes from the medical and hospital fields, pharmaceuticals, and down the list genetics research and academic institutions. I have no evidence, but I suspect the volume of direct-to-consumer sales of tests for genealogy the past 24 months has taken both Illumina and Thermo Fisher by surprise. Probably not enough to be much more than an interesting blip among their biggest revenue sectors, but interesting nonetheless.
Helix.com is a bit of a strange business model...really more of a DNA "app store" than anything else. And genealogy really ain't in the mix. None of Helix's investors or equity partners have any stake in genealogy. It's all about health and pharmaceuticals. The money behind Helix? Illumina, the Mayo Clinic, DFJ Growth, Kleiner Perkins Caufield Byers, Sutter Hill Ventures, and Warburg Pincus.
Absoposilutely nothing wrong with DNA testing for health, fitness, nutrition, medical, or pharmaceutical reasons. I simply have little experience with DNA testing for those purposes. Fifty bucks for a report to help you choose the kinds of wines that would be tastiest to you? Really?
But let's address Levi's note about Helix stating they test 100x more data than typical at-home tests. Is that statement true? Yes; it is. But the fact that it's massively difficult to find, on their website, what you actually get for your purchase should be the first red flag.
Here's the deal. Helix is providing NextGen, full-sequence testing. But what they're running is what's known as a WES, a Whole-Exome Sequence. The exome consists of the DNA that comprises all of your known protein-coding genes. A WES test is looking at a whopping 75-100 times more places on your chromosomes than do the tests from AncestryDNA, Family Tree DNA, 23andMe, MyHeritage, and Living DNA.
More is better, right? Well...not necessarily. See, if stuff changes, mutates, in your protein-encoding genes, it has the potential to spell survivability issues for the organism. What population geneticists and genealogists are most interested in are places in the genome that show variability without killing off the species. Always a good thing, that.
These are SNPs and STRs. STRs I won't address (because neither the typical DTC test nor Helix looks at them) and this will be only a thumbnail snapshot of SNPs. Pronounced "snip," these are specific places in the chromosomes where consistent, identified population differences have been found. Technically, for a SNV (single nucleotide variation) to become elevated to SNP (single nucleotide polymorphism) status, it has to be found in a significant amount of the general population. Most take that as meaning greater than 1% of the general population; no serious genetic condition occurs with that frequency.
SNPs come in basically two major flavors: linked and causative. Linked SNPs aren't found within our genes and don't impact protein productions or functions. Though they can show a correlation to proclivity for a certain disease or a person's reaction/tolerance to certain drugs, linked SNPs are free to mutate without endangering the organism. Most SNPs are of this "linked" variety.
Causative SNPs are located within the chromosomal regions of genes and can be either coding or non-coding. Coding SNPs change the actual amino acids of the gene's protein product; non-coding SNPs are found in a gene's regulatory region and, while they don't change the amino acid sequences produced, they can change the location, timing, or level of gene expression.
But despite a seemingly straightforward definition, knowing precisely how many human SNPs are cataloged at any given time is like nailing Jello to a tree: depends on who you ask, how they're measuring it, and what data they're looking at. I've been writing recently that they're are around 10-12 million SNPs. I was called on that just a couple of days ago by none other than the esteemed Louis Kessler, who said there were about 20 million. My number was drawn from what's stated by the
National Institutes of Health and the
National Center for Biotechnology Information. There are more--and fewer--SNPs cataloged in different places than 10 million.
SNPedia currently shows 109,677 in their database, and in
October 2015, writing in Nature magazine, the 1000 Genomes Project Consortium said that they had characterized "84.7 million single nucleotide polymorphisms (SNPs), 3.6 million short insertions/deletions (indels), and 60,000 structural variants..." So the NIH's estimate of around 10 million is probably serviceable for our purposes.
I believe--but don't hold me to it; we'd need our resident GEDmatch authority, Aaron Wells, to set us straight--that the GEDmatch database contains a little under 2 million SNPs. Our current, popular DTC genealogy tests look at, maximally, about 702,000 SNPs for FTDNA and MyHeritage on the Illumina OmniExpress chip; 630,000 SNPs on the 23andMe GSA chip; and the verdict is still out on the total that is being used by the Thermo Fisher Scientific chip newly deployed by Living DNA. The 702K and 630K counts overlap only by about 22%; in other words, there are only about 150K in-common SNPs tested...call it 1.1 million aggregate SNPs now tested for genealogy.
Back to Helix and the "100x more data than typical at-home tests." Remember the whole
exome sequencing, not whole
genome sequencing? The test does examine a vastly greater number of markers than our usual tests...but it's looking primarily within the regions of protein encoding genes; it isn't looking at those "linked" SNPs that predominate and that don't affect the work the genes do. I'm not certain we have precise numbers, but
the best data I know of comes from ISOGG and shows that the Helix test--of those 50 million or so markers it examines--will look at about 144K of the same SNPs as FTDNA, and only about 45K of the same as 23andMe, only 7%.
So it comes down to the question, "Are those 50 million markers ones I care about?" If you're testing for health purposes, yes; absolutely. If for genealogy/ancestry purposes...meh; not so much. This is the reason I've never been particularly interested in Helix, but recently jumped at the chance to get a $200 whole-genome 30X sequencing (WGS as opposed to WES). That same paper in Nature linked earlier states, "We find that a typical genome differs from the reference human genome at 4.1 million to 5.0 million sites [and that] >99.9% of variants consist of SNPs and short indels [insertions/deletions]." For me, that's the increased coverage I'm interested in and I want to see the massive data file that results from whole-genome sequencing. The Helix test looks at very few of those potential 5 million variants and, even at 50 million loci tested, only at about 1.6% of your whole genome.
Regardless, neither a WGS or WES test can, today, be used for genealogical comparison or matching. Research thoroughly based upon your intended result. But for genealogy, we're at the beginning of the big holiday sales season, and I'd consider looking at one of the more popular DNA tests and leaving Helix on the shelf, at least for now.
