My (new!) cousin is trying to upload her AncestryDNA results to ftDNA, and running into errors.  Two questions: * Is ftDNA accepting AncestryDNA's results from its most current chip? * Does she upload the zipped or unzipped file? Really, I looked on ftDNA and DNAADOPTION.com.  Not finding an answer to the latter question. Thanks!

This problem has been reported many places. Apparently FTDNA is looking for a specific line of text in the header, and AncestryDNA changed that. The work-around was to open the file with a text editor (e.g. Windows WordPad) and change the phrase "AncestryDNA array version: V1.0" to read "AncestryDNA array version: V2.0". But now people are reporting that AncestryDNA resurrected the V2.0 phrase for fresh downloads and something isn't working again. The saga continues. MyHeritage uploads have also been a problem for months. There is a thread on the FTDNA forums which you could check from time to time for the latest news. http://forums.familytreedna.com/showthread.php?p=443198#post443198 FTDNA will accept either the zipped or unzipped file. Ann Turner On Sun, Aug 27, 2017 at 11:12 AM, Brooks Family <[email protected]> wrote: > My (new!) cousin is trying to upload her AncestryDNA results to ftDNA, and > running into errors. Two questions: > > * Is ftDNA accepting AncestryDNA's results from its most current chip? > > * Does she upload the zipped or unzipped file? > > Really, I looked on ftDNA and DNAADOPTION.com. Not finding an answer to > the latter question. > > Thanks! > > > ------------------------------- > To unsubscribe from the list, please send an email to > [email protected] with the word 'unsubscribe' without > the quotes in the subject and the body of the message

I haven't heard any problems with people uploading AncestryDNA data from the current chip. Uploading the zipped file should work ok. If there are problems, I would open the raw data file and make sure that it looks complete. The zipped file should be about 5.7 MB in size. Tim Janzen -----Original Message----- From: GENEALOGY-DNA [mailto:[email protected]] On Behalf Of Brooks Family Sent: Sunday, August 27, 2017 11:12 AM To: [email protected] Subject: [DNA] Uploading AncestryDNA results to ftDNA My (new!) cousin is trying to upload her AncestryDNA results to ftDNA, and running into errors. Two questions: * Is ftDNA accepting AncestryDNA's results from its most current chip? * Does she upload the zipped or unzipped file? Really, I looked on ftDNA and DNAADOPTION.com. Not finding an answer to the latter question. Thanks!

Hello everyone, if anyone has a match on both paternal and maternal side, would you be so kind to help me? I’m trying to figure out how 23andMe is showing this in their data. Thanks in advance, feel free to respond either directly to my email "ahnen (at) awest (dot) de” or reply to the list Andreas Andreas West Meine Vorfahren / my ancestors (8 generations): http://www.wikitree.com/treewidget/Basso-23/5 <http://www.wikitree.com/treewidget/Basso-23/5> Author of https://www.dnagenealogy.tools <http://dnagenealogy.tools/>

Does anyone have any idea what kind of a formula could be used to compare the “likely closeness of the relationship (on average)” of two “DNA cousins” who share a number of segments? 23andMe, for instance, ranks more highly someone who shares 40 cM in 3 HIRs than someone else who shares 45 cM in 2 HIRs. Best, Eric <https://keyserver.pgp.com/vkd/DownloadKey.event?keyid=0xE0F58E0F1AF7E6F2> OpenPGP: 0x1AF7E6F2 ● Skype: oneota ● XMPP/OTR: <mailto:[email protected]> [email protected] ● Silent Circle: +1 312 614-0159

I'm confused, nothing new there. Tim, are you recommending that "all serious autosomal genetic genealogists test at 23andMe"? So then, if you upload your 23and Me results to FTDNA, where you have already tested, or to Gedmatch, which already has your Family Finder results, are there any conflicts? One person, 2 sets of results? Bill -----Original Message----- From: GENEALOGY-DNA [mailto:[email protected]] On Behalf Of Tim Janzen Sent: Friday, 25 August 2017 10:35 AM To: [email protected] Subject: Re: [DNA] schism Dear All, I thought I would let you know that over the past week or so I have been running comparisons between the 23andMe v5 data and the data from the other companies to see how many autosomal SNPs overlap between the datasets. See the top chart at https://isogg.org/wiki/Autosomal_SNP_comparison_chart#Autosomal_SNPs. I have several more Genes for Good comparisons to run and then this chart will be complete. My initial impression is that 23andMe is doing a reasonably good job with imputation for the matching segment data for DNA Relatives. My impression is that the imputation for the Genesis Project at GEDmatch is not quite as good so far. It is unfortunate that this "schism" exists for 23andMe v5 and Living DNA data, but that is the way that it is. I wonder if and/or when Family Tree DNA will be converting over to the Illumina Global SNP Array chip. My suggestion at this point is that it would be prudent to test all of your closest family members on a GSA chip (23andMe v5 or Living DNA) so that you can upload the data to GEDmatch and be assured you will being seeing all of your matches who have tested only on a GSA chip who share reasonably long (say 15cMs to 30 cMs) DNA segments with your family or you. Mike, in my opinion all serious autosomal genetic genealogists test at 23andMe. 23andMe has well over 2 million people in their database so it is would be unwise for a serious autosomal genetic genealogist to not test there. 23andMe certainly has the best DNA comparison tools relative to Ancestry.com, MyHeritage and FTDNA's Family Finder. GEDmatch offers unsurpassed genetic comparison features that we need as well. Sincerely, Tim Janzen -----Original Message----- From: GENEALOGY-DNA [mailto:[email protected]] On Behalf Of Michael Fisher Sent: Tuesday, August 22, 2017 7:33 AM To: [email protected] Subject: Re: [DNA] schism Hi I was thinking of testing at 23andMe already having AncestryDNA and FTDNA but will give it a miss now. How does LivingDNA compare ? How do these new tests perform in the terms of matches to each other with the same provider or to a similar new type test provider ? Mike Fisher ------------------------------- To unsubscribe from the list, please send an email to [email protected] with the word 'unsubscribe' without the quotes in the subject and the body of the message

I do have matches who match both my paternal and maternal sides, but it's not through 23AndMe -----Original Message----- From: Andreas West <[email protected]> To: genealogy-dna <[email protected]>; DNA-NEWBIE <[email protected]> Sent: Sat, Aug 26, 2017 2:38 am Subject: [DNA] Looking for someone who has matches on both maternal and paternal side Hello everyone,if anyone has a match on both paternal and maternal side, would you be so kind to help me? I’m trying to figure out how 23andMe is showing this in their data.Thanks in advance, feel free to respond either directly to my email "ahnen (at) awest (dot) de” or reply to the listAndreas Andreas WestMeine Vorfahren / my ancestors (8 generations): http://www.wikitree.com/treewidget/Basso-23/5 <http://www.wikitree.com/treewidget/Basso-23/5>Author of https://www.dnagenealogy.tools <http://dnagenealogy.tools/> -------------------------------To unsubscribe from the list, please send an email to [email protected] with the word 'unsubscribe' without the quotes in the subject and the body of the message

Bill, This happens sometimes and it is not an issue. I've tested, with an original saliva sample at FTDNA, 23andMe, AncestryDNA, MyHeratige, LivingDNA, and Genos. I got an account at each company. I uploaded the first 4 to GEDmatch and got a different Kit ID each time - it looks like 4 different people with the same email. Others have done this too. Some get confused or frustrated and upload the same kit again - I've seen up to 4 kits from the same AncestryDNA download (each got a different kit ID). It's OK to do this, but anyone who does, should make all but one kit "Private". You are not permitted to upload an atDNA kit at FTDNA if you already have an atDNA kit there - only one atDNA kit per customer. So there are no conflicts. Sometimes it is confusing when you get kits from GEDmatch from the same email as a Match more than once - is it the same person, or parent child, or what (sometimes they use the same name, sometimes its altered slightly, sometimes a lot). Since I Triangulate my Matches, I'm comparing them to each other. If they share about 3600cM that's usually a parent/child. I then rerun the comparison with the graphics on - a parent/child will have some differences, a duplicate kit will be solid blue. Hope this helps. Sent from my iPhone > On Aug 26, 2017, at 12:46 AM, Bill Webster <[email protected]> wrote: > > I'm confused, nothing new there. > Tim, are you recommending that "all serious autosomal genetic genealogists > test at 23andMe"? > So then, if you upload your 23and Me results to FTDNA, where you have > already tested, or to Gedmatch, which already has your Family Finder > results, are there any conflicts? One person, 2 sets of results? > Bill

Andreas, Since both my husband and I have tested at 23andme, our daughter's matches are shown m for maternal and p for paternal and m/p if the match is on both the maternal and paternal side. Marilyn On Sat, Aug 26, 2017 at 5:38 AM, Andreas West <[email protected]> wrote: > Hello everyone, > > > if anyone has a match on both paternal and maternal side, would you be so > kind to help me? I’m trying to figure out how 23andMe is showing this in > their data. > > > Thanks in advance, feel free to respond either directly to my email "ahnen > (at) awest (dot) de” or reply to the list > > > Andreas > > Andreas West > Meine Vorfahren / my ancestors (8 generations): http://www.wikitree.com/ > treewidget/Basso-23/5 <http://www.wikitree.com/treewidget/Basso-23/5> > Author of https://www.dnagenealogy.tools <http://dnagenealogy.tools/> > > ------------------------------- > To unsubscribe from the list, please send an email to > [email protected] with the word 'unsubscribe' without > the quotes in the subject and the body of the message

There are many different combinations of total cM and number of segments found for a given relationship, as shown in the scattergram from this blog post: http://blog.23andme.com/news/announcements/how-many-relatives-do-you-have/ The blobs overlap, so the colors aren't clear, but if you could plot those values on the scattergram, you'd probably find that the three-segment match fell closer to the higher relationship. It's very important to note that 23andMe predicts a range of relationships -- one might be classified as say a third to fifth cousin, and the other a third to six cousin. The total cM is the single best predictor, and it's the easiest to handle computationally. Ann Turner On Sat, Aug 26, 2017 at 12:06 AM, Eric S Johnson <[email protected]> wrote: > Does anyone have any idea what kind of a formula could be used to compare > the “likely closeness of the relationship (on average)” of two “DNA > cousins” who share a number of segments? > > > > 23andMe, for instance, ranks more highly someone who shares 40 cM in 3 HIRs > than someone else who shares 45 cM in 2 HIRs. > > > > Best, > > Eric > > <https://keyserver.pgp.com/vkd/DownloadKey.event?keyid=0xE0F58E0F1AF7E6F2 > > > OpenPGP: 0x1AF7E6F2 ● Skype: oneota ● XMPP/OTR: > <mailto:[email protected]> [email protected] ● Silent Circle: +1 > 312 614-0159 > > > > > ------------------------------- > To unsubscribe from the list, please send an email to > [email protected] with the word 'unsubscribe' without > the quotes in the subject and the body of the message >

Dear Eric, There are some probabilities here, but my suggestion is that you only use a "formula" for people who share more than about 75 cMs with you. Such people are likely to be 3rd cousins or closer. For people who share about 50 cM to 75 cMs, the probabilities favor 3rd cousins once removed to 2nd cousins once removed. Below 50 cMs, I categorize them all as "distant" cousins to 3rd cousins until I can better sort out the exact relationship using either genetic analysis or from genealogical research. I use cousin clusters and run comparisons to determine how much DNA a new match shares on average with the appropriate cousin cluster. This allows me to be more precise in terms of predicting the true genealogical relationship than I can be if I only do a one-to-one comparison. Sincerely, Tim Janzen -----Original Message----- From: GENEALOGY-DNA [mailto:[email protected]] On Behalf Of Eric S Johnson Sent: Saturday, August 26, 2017 12:06 AM To: [email protected] Subject: [DNA] cM v. segments Does anyone have any idea what kind of a formula could be used to compare the "likely closeness of the relationship (on average)" of two "DNA cousins" who share a number of segments? 23andMe, for instance, ranks more highly someone who shares 40 cM in 3 HIRs than someone else who shares 45 cM in 2 HIRs. Best, Eric

Dear Bill, Yes, I am recommending that all serious autosomal genetic genealogists test at 23andMe. You simply can't afford to miss any of your high quality matches and you will get a lot of them at 23andMe. You won't necessarily have an easy time getting the pedigree charts from all of your matches there, but sometimes you can figure out a match's pedigree chart if they have an unusual name even if they don't do anything other than opt in to open sharing. Keep in mind that to do optimal triangulation, you need the matching segment data AND the pedigree charts from as many people as possible. You only need to have one set of results in Family Finder. I think that it makes sense for the serious autosomal genetic genealogist to do the Family Finder test for all of their close family members because testing at FTDNA gives you the full set of SNP data from the Omni Express chip. This optimizes the probability that the matching segment data will be consistent the most of the other people in the Family Finder dataset. What we don't know at this point is if FTDNA will eventually be moving over to the GSA chip. In terms of uploading to GEDmatch, I would upload a 23andMe version 3 file if you have it since that dataset has the largest number of autosomal SNPs. Otherwise, I would suggest uploading a Family Finder and/or an Ancestry.com dataset to GEDmatch. If you do the 23andMe version 5 test then you will upload to the GEDmatch Genesis project where your results will be compared to other 23andMe version 5 kits and to Living DNA kits. If you already have a Family Finder kit in the regular GEDmatch, then you can simply leave it alone. You won't upload 23andMe version 5 test results into Family Finder unless FTDNA switches to the GSA chip, at which point in time FTDNA might allow that. Sincerely, Tim -----Original Message----- From: GENEALOGY-DNA [mailto:[email protected]] On Behalf Of Bill Webster Sent: Friday, August 25, 2017 9:46 PM To: [email protected] Subject: Re: [DNA] schism I'm confused, nothing new there. Tim, are you recommending that "all serious autosomal genetic genealogists test at 23andMe"? So then, if you upload your 23and Me results to FTDNA, where you have already tested, or to Gedmatch, which already has your Family Finder results, are there any conflicts? One person, 2 sets of results? Bill

Thanks very much for the chart and the extensive work that went into it. The low point on the chart is the cell-pair for 23andMe v2 and 23andMe v5. So is 23andMe going to upgrade all of their prior testers' results to v5? Otherwise, they are diluting the value of an earlier test at 23andMe and thus diluting the value of the size of their database and the value of their tools. The only way to make an apples to apples comparison on 23andMe is to make sure every test is an apple. Are they going to do that? ---------------------------------------------------------------------- Date: Thu, 24 Aug 2017 17:35:00 -0700 From: "Tim Janzen" <[email protected]> Subject: Re: [DNA] schism Dear All, I thought I would let you know that over the past week or so I have been running comparisons between the 23andMe v5 data and the data from the other companies to see how many autosomal SNPs overlap between the datasets.  See the top chart at https://isogg.org/wiki/Autosomal_SNP_comparison_chart#Autosomal_SNPs. I have several more Genes for Good comparisons to run and then this chart will be complete.  My initial impression is that 23andMe is doing a reasonably good job with imputation for the matching segment data for DNA Relatives. My impression is that the imputation for the Genesis Project at GEDmatch is not quite as good so far.  It is unfortunate that this "schism" exists for 23andMe v5 and Living DNA data, but that is the way that it is.  I wonder if and/or when Family Tree DNA will be converting over to the Illumina Global SNP Array chip.  My suggestion at this point is that it would be prudent to test all of your closest family members on a GSA chip (23andMe v5 or Living DNA) so that you can upload the data to GEDmatch and be assured you will being seeing all of your matches who have tested only on a GSA chip who share reasonably long (say 15cMs to 30 cMs) DNA segments with your family or you.  Mike, in my opinion all serious autosomal genetic genealogists test at 23andMe.  23andMe has well over 2 million people in their database so it is would be unwise for a serious autosomal genetic genealogist to not test there.  23andMe certainly has the best DNA comparison tools relative to Ancestry.com, MyHeritage and FTDNA's Family Finder.  GEDmatch offers unsurpassed genetic comparison features that we need as well. Sincerely, Tim Janzen  -----Original Message----- From: GENEALOGY-DNA [mailto:[email protected]] On Behalf Of Michael Fisher Sent: Tuesday, August 22, 2017 7:33 AM To: [email protected] Subject: Re: [DNA] schism Hi I was thinking of testing at 23andMe already having AncestryDNA and FTDNA but will give it a miss now. How does LivingDNA compare ? How do these new tests perform in the terms of matches to each other with the same provider or to a similar new type test provider ? Mike Fisher

I'm in the process of doing some detailed comparisons of the SNP sets. It's just not the sheer number of SNPs that counts, but how informative they are for most customers. For instance, 421 of the v4 SNPs have fewer than 10 instances in the GenBank database of 37,545 records, while only 25 v5 SNPs are in that category. Hopefully some of those "wasted" SNPs were replaced with more informative SNPs. Of course, someone with a rare SNP would love to know about it, but chip-based testing is no substitute for full mitochondrial sequencing. Ann Turner On Tue, Aug 22, 2017 at 4:13 AM, Ian Logan <[email protected]> wrote: > Eric > > Whilst we are talking about the new V5 - may I just mention the mtDNA SNPs. > > Unfortunately, it would seem V5 is not as good as V4 for mtDNA analysis; > and 23andMe are currently only reporting on 14.7% of mtDNA bases > (whereas the V4 chip results very usefully reported on 19.7%) > > And to make things worse, there appears to be around an 8.3% 'no call' > level; > which knocks the coverage back even further. > > Overall, 23andMe appear to have missed the opportunity to improve their > mtDNA reporting - and in fact seem to have gone backwards. > > I hope future analysis of the Autosomal results do not prove to disappoint > also. > > Ian > --------------------- >

23andMe will use imputation when comparing v5 results with prior versions. Imputation is the prediction of the results for an untested SNP, based on the results for nearby tested SNPs. I have done some experiments with imputed data I obtained from DNA.Land, and it is about 99% accurate for the SNPs on chips. That 1% difference can potentially affect relationship predictions, though, so 23andMe is developing proprietary algorithms to handle discrepancies. Also, DNA.Land imputation of 39,000,000 SNPs is based on whole genome sequencing in 1000 Genomes Project. 23andMe will have a much larger reference set in their internal database, so they may be able to achieve higher accuracy. I am going to get a v5 test for myself to compare results with prior versions, so I will have some head-to-head data in a few weeks. Ann Turner On Fri, Aug 25, 2017 at 3:52 AM, Wesley Johnston <[email protected]> wrote: > > > Thanks very much for the chart and the extensive work that went into it. > The low point on the chart is the cell-pair for 23andMe v2 and 23andMe v5. > So is 23andMe going to upgrade all of their prior testers' results to v5? > Otherwise, they are diluting the value of an earlier test at 23andMe and > thus diluting the value of the size of their database and the value of > their tools. The only way to make an apples to apples comparison on 23andMe > is to make sure every test is an apple. Are they going to do that? > > ---------------------------------------------------------------------- > > Date: Thu, 24 Aug 2017 17:35:00 -0700 > From: "Tim Janzen" <[email protected]> > Subject: Re: [DNA] schism > > Dear All, > I thought I would let you know that over the past week or so I have been > running comparisons between the 23andMe v5 data and the data from the other > companies to see how many autosomal SNPs overlap between the datasets. See > the top chart at > https://isogg.org/wiki/Autosomal_SNP_comparison_chart#Autosomal_SNPs. I > have several more Genes for Good comparisons to run and then this chart > will > be complete. My initial impression is that 23andMe is doing a reasonably > good job with imputation for the matching segment data for DNA Relatives. > My impression is that the imputation for the Genesis Project at GEDmatch is > not quite as good so far. It is unfortunate that this "schism" exists for > 23andMe v5 and Living DNA data, but that is the way that it is. I wonder > if > and/or when Family Tree DNA will be converting over to the Illumina Global > SNP Array chip. My suggestion at this point is that it would be prudent to > test all of your closest family members on a GSA chip (23andMe v5 or Living > DNA) so that you can upload the data to GEDmatch and be assured you will > being seeing all of your matches who have tested only on a GSA chip who > share reasonably long (say 15cMs to 30 cMs) DNA segments with your family > or > you. > > Mike, in my opinion all serious autosomal genetic genealogists test at > 23andMe. 23andMe has well over 2 million people in their database so it is > would be unwise for a serious autosomal genetic genealogist to not test > there. 23andMe certainly has the best DNA comparison tools relative to > Ancestry.com, MyHeritage and FTDNA's Family Finder. GEDmatch offers > unsurpassed genetic comparison features that we need as well. > Sincerely, > Tim Janzen > > -----Original Message----- > From: GENEALOGY-DNA [mailto:[email protected]] On Behalf > Of > Michael Fisher > Sent: Tuesday, August 22, 2017 7:33 AM > To: [email protected] > Subject: Re: [DNA] schism > > Hi > > I was thinking of testing at 23andMe already having AncestryDNA and > FTDNA but will give it a miss now. > > How does LivingDNA compare ? > > How do these new tests perform in the terms of matches to each other > with the same provider or to a similar new type test provider ? > > Mike Fisher > > > > ------------------------------- > To unsubscribe from the list, please send an email to > [email protected] with the word 'unsubscribe' without > the quotes in the subject and the body of the message >

Dear All, I thought I would let you know that over the past week or so I have been running comparisons between the 23andMe v5 data and the data from the other companies to see how many autosomal SNPs overlap between the datasets. See the top chart at https://isogg.org/wiki/Autosomal_SNP_comparison_chart#Autosomal_SNPs. I have several more Genes for Good comparisons to run and then this chart will be complete. My initial impression is that 23andMe is doing a reasonably good job with imputation for the matching segment data for DNA Relatives. My impression is that the imputation for the Genesis Project at GEDmatch is not quite as good so far. It is unfortunate that this "schism" exists for 23andMe v5 and Living DNA data, but that is the way that it is. I wonder if and/or when Family Tree DNA will be converting over to the Illumina Global SNP Array chip. My suggestion at this point is that it would be prudent to test all of your closest family members on a GSA chip (23andMe v5 or Living DNA) so that you can upload the data to GEDmatch and be assured you will being seeing all of your matches who have tested only on a GSA chip who share reasonably long (say 15cMs to 30 cMs) DNA segments with your family or you. Mike, in my opinion all serious autosomal genetic genealogists test at 23andMe. 23andMe has well over 2 million people in their database so it is would be unwise for a serious autosomal genetic genealogist to not test there. 23andMe certainly has the best DNA comparison tools relative to Ancestry.com, MyHeritage and FTDNA's Family Finder. GEDmatch offers unsurpassed genetic comparison features that we need as well. Sincerely, Tim Janzen -----Original Message----- From: GENEALOGY-DNA [mailto:[email protected]] On Behalf Of Michael Fisher Sent: Tuesday, August 22, 2017 7:33 AM To: [email protected] Subject: Re: [DNA] schism Hi I was thinking of testing at 23andMe already having AncestryDNA and FTDNA but will give it a miss now. How does LivingDNA compare ? How do these new tests perform in the terms of matches to each other with the same provider or to a similar new type test provider ? Mike Fisher

GEDmatch has this announcement at the top of my login page: “Aug 10 2017 NOTICE to 23andMe Customers: 23andMe is now using the GSA chip for their new V5 raw DNA file results. This format is not compatible with the regular GEDmatch upload, but can be used with the GEDmatch Genesis upload. Use the link in the lower right column of this page.” Is our ability to “compare everyone with everyone” on GEDmatch now broken? I.e. all new 23andMe kits can’t be compared to any of the tens of thousands of existing GEDmatch kits? Genesis says “GEDmatch Genesis database will be separate from the main GEDmatch database, and comparisons for one will not show entries made in the other. Eventually, the 2 databases will be merged, and results will include entries from both” … but I wonder how long it’ll take for the merge; and it sounds like, for the moment, the schism is real ☹ Best, Eric <https://keyserver.pgp.com/vkd/DownloadKey.event?keyid=0xE0F58E0F1AF7E6F2> OpenPGP: 0x1AF7E6F2 ● Skype: oneota ● XMPP/OTR: <mailto:[email protected]> [email protected] ● Silent Circle: +1 312 614-0159

Hi I was thinking of testing at 23andMe already having AncestryDNA and FTDNA but will give it a miss now. How does LivingDNA compare ? How do these new tests perform in the terms of matches to each other with the same provider or to a similar new type test provider ? Mike Fisher

23andMe will revise its algorithm for backward compatibility to its own previous versions, based on observations of its large internal database. That's not something GEDmatch can do, right now anyway. Ann Turner On Tue, Aug 22, 2017 at 12:51 PM, Brooks Family <[email protected]> wrote: But what about databases? 23andme seems to have truly unique database of > individuals to compare against. > > And the in-common function that they added earlier this year is a real > bonus when dealing with silent cousins. >

But what about databases?  23andme seems to have truly unique database of individuals to compare against. And the in-common function that they added earlier this year is a real bonus when dealing with silent cousins. On 8/22/17 9:35 AM, [email protected] wrote: > LivingDNA and 23andMe v4 are very similar, using the same base GSA chip of > ~600K SNPs but with their own custom SNPs added for various reasons. > Segment matching across those two platforms should be roughly comparable to > comparing two Illumina 700K chips and better than comparing say FTDNA vs > AncestryDNA v2 or 23andMe v4. The GSA chip is probably the wave of the > future; LivingDNA has said that they expected the 700K chip to be phased > out at some point, so they chose to use the GSA chip from the get-go.