ISOGG, https://isogg.org, has updated its autosomal SNP comparison chart: https://isogg.org/wiki/Autosomal_SNP_comparison_chart. The chart compares the following DNA companies and chip sets: AncestryDNA (v1 and v2 chips); FamilyTree DNA (FTDNA), 23andMe (v2, v3, v4, and v5 chips), Genes for Good, LivingDNA, MyHeritageDNA, and Genographic (Geno)/ Helix (not sure if it includes the FTDNA kits for Geno as they have the same number of SNPs as the Geno Helix kits, but haven’t checked to see if both use the same SNPs).
If you are transferring raw data from one company to another or using GEDmatch/GEDmatch Genesis, it’s important to realize the differences in SNPs as they can cause two things to happen. First, your matches may be off in terms of relationship. For example a full sibling could show as a half sibling if the number of SNPs compared vary significantly because Sibling A tested at one company and Sibling B tested at another company where the SNP differences were large enough to affect the relationship estimate. Next, you may lose valid DNA matches for the same reason. If you look at the ISOGG chart in the link above, at the low end the differences are only a few thousand SNPs, but at the high end, the differences can be off by almost 600,000 which is a huge difference. Many of the differences are in the 200.,000 to 400,000 SNP range.
I regularly use the ISOGG Wiki for different things, including the above link as well as https://isogg.org/wiki/Autosomal_DNA_testing_comparison_chart and https://isogg.org/wiki/List_of_DNA_testing_companies (this was one of several sites I am using to create an improved and updated Links Section https://upsdownsfamilyhistory.wordpress.com/links/) which I hope to roll out within the next couple of weeks.
It’s slow going on updating the Links Section as I had to dig through numerous websites, blogs, Android apps, and iTunes apps. I will be adding it as a separate blog post initially and later replacing the text in the Links Section with what’s in the new blog post. I started the update back in late July and work on it as time allows. I tentatively have it set as a scheduled post for October 31, 2018, but hope to have it done much sooner than that. The blog post is almost ready, except for going through and making sure it has been tidied up.