Ancient I-M253 samples list

[spruithean]

Y6626 is part of L1237, which is downstream of Z63, which is normally dubbed the "Eastern Germanic" branch. Not L338 it seems.

https://www.yfull.com/tree/I-L1237/, one of Collegno Longobards is I-Y2245, which is just upstream of L1237.

[JonikW]

Y6626 is part of L1237, which is downstream of Z63, which is normally dubbed the "Eastern Germanic" branch. Not L338 it seems.

https://www.yfull.com/tree/I-L1237/, one of Collegno Longobards is I-Y2245, which is just upstream of L1237.

Thanks spruithean, knew I could rely on you.

[deadly77]

Just to share some Polish ancient I1...

This was originally posted by Tomenable on another forum



Credit goes to Tomenable.

Perhaps some of these could go to the ancient I1 map at some point, perhaps if they are fully vetted, or tentatively added for now? I should add that the Kowalewko samples have been attributed to the Wielbark culture (AFAIK) which is where the early Goths, Gepids and Rugii were believed to have originated before migrating onward to the Chernyakhov cultural sphere.

Yes, these two have been on radar for a while - they've been on this forum since post #3. They keep getting brought up every so often but I've yet to find any real data on these samples.

I mentioned the issues that I have had trying to get reliable data to analyze or any more information aside from a poster on academia earlier on on this thread here https://anthrogenica.com/showthread....l=1#post537149 and https://anthrogenica.com/showthread....l=1#post559825

Unless something changes in terms of available data that I'm able to analyze or more information, there's nothing I can do to verify the SNP calls in those samples (or any of the other unnamed I1). So far, no paper published or update to data at ENA that I'm aware of.

I have tried take BAM files of other samples from this thread that I can run succesfully with IGV and run those through Felix Immanuel's BAM analysis kit, which seemed to be how folks were analyzing these samples in another thread on AG (after converting from FASTQ using one of Felix's other tools). Felix's BAM analysis kit hasn't been updated in a long time (2015?) and seems to really struggle with samples that have a lot of gaps in coverage. I often get very different results to when I do a manual review of the BAM. The Felix BAM analysis kit misses out a lot of SNPs and reports some that aren't there when looking at the BAM file via IGV. I don't really trust this tool for ancient sample analysis.

I'm going to hold off adding these to my map until the authors publish a paper that has more detail than a poster abstract or I can find a BAM file that can be read and verified in IGV. I'm happy to do the analysis, but from my perspective, there is nothing currently that can be reliably analyzed. If you're able to find anything, let me know.

[spruithean]

Understandable, and this has been my concern as well in regard to which samples can be mapped. I'm surprised that we still haven't heard much about these Polish results since the small bits of info were first released, it's disappointing, but I bet eventually something will finally be released.

[deadly77]

Understandable, and this has been my concern as well in regard to which samples can be mapped. I'm surprised that we still haven't heard much about these Polish results since the small bits of info were first released, it's disappointing, but I bet eventually something will finally be released.

Yeah, I tried looking in the "Early Medieval aDNA from Poland Coming Soon" thread on Anthrogenica but over 244 pages comprising 2443 posts, it's difficult to find any reference to actual data buried amongst the conjecture and bickering. Perhaps you could try asking Tomenable in the other forum .

[deadly77]

Yeah, I tried looking in the "Early Medieval aDNA from Poland Coming Soon" thread on Anthrogenica but over 244 pages comprising 2443 posts, it's difficult to find any reference to actual data buried amongst the conjecture and bickering. Perhaps you could try asking Tomenable in the other forum .

And the haplogroup and subclade assignments for these two samples may very well be correct. But after seeing so many misassignments in publications and secondary analysis of I1 samples - for example: SVK-A1 listed as I1b in the Ebenesersdóttir et al. paper; CL63 listed as I1-Z79 in the Amorim et al paper and numerous holes in the supplementary information regarding calls on that sample and SZ45; Carlos Quiles misassigning HSJ-A1 and SSJ-A2 to I-A8182, SBT-A1 to I-S26062 and repeating the incorrect assignment in the paper of SVK-A1 as I1b, etc.; it's important to check against the raw data for I1 ancient DNA (and indeed all haplogroups).

[deadly77]

That's pretty frustrating that Mr. Quiles would do it that way but I suppose he's not alone, and it's pretty disappointing that the teams working on these ancient DNA works would be so neglectful of the fact that the ISOGG nomenclature and alphabet soup change every year, the only thing that doesn't change is the shorthand nomenclature. I-L338 is always I-L338, this way confusion is very low. I can't even look at older mentions (or newer mentions) of R1b and R1a clades if they don't have their terminal SNP mentioned. The size of some of those alphabet soups for R1 clades is ridiculous

Keep us posted on those urm guys, they could get really interesting and even a tentative placement as simply I-DF29 for both is a hell of lot better than "ND" and "BCDEF". I was reading (yet again) the Sigtuna paper yesterday, quite an interesting read in terms of where some of these people likely came from in relation to Sigtuna.

Thanks again for all the work you do on these BAM files, it's definitely not something I'm able to do, or probably capable of

So Hunter refers to YFulls data? I had wondered as there is another phylogeography tracker that seems based off of FTDNA's data, which the creator of that site has begun posting here. http://scaledinnovation.com/gg/snpTracker.html , For my SNP, Hunter's tracker and the above tracker are a fair bit different in terms of dating and I suppose naming. YFull still has A14097 lumped in with A14094 (my fault really as I've been tardy with anything related to sending off data to them), while the other has A14097 above A14094/M4050. Not too big of a deal though.

I had suspected that AE sample to be a typo. Seemed to similar in the beginning to AED.



I think Razib Khan was simply saying that I1 carriers hitched a ride with these R1b rich groups wherever it may have been and "prospered" off the fortune of these R1b clans. This could have happened anywhere, and given the proliferation of I1 probably occured in the Bronze Age it could have happened in Northern Europe? I didn't quite get the impression that he was implying (totally) I1 literally hitched a migratory ride somewhere with the R1b folks, more that it was just benefit by association. Perhaps a bit parasitic...

I wish we had more definitive answers, I1 is a missing link in a lot of the ancient DNA...



Exactly this, which sort of fits with the definition, (stay with me here my brain is not cooperating today), I1's TMRCA is sometime in the Bronze Age if the YFull and other TMRCA estimates are correct, that places it sometime in 27th-25th century BC if not earlier or later (some context, the Instructions of Shuruppak were likely written sometime in the space of 2600 to 2500 BCE). So I1 came out of hiding at some point around this time, yet we've found potential I1 samples pre-dating this by a long time as BAB5 is dated to around 5600-4900 BCE, and BAL051 is dated even earlier than that. However their SNP calls are rather underwhelming for being definitive for being I1-M253. This is why deadly and several others have begun to label these samples as pre-I1 (including the Nordic LN and BA guys), but we already know that I1's split from I2 is likely quite some time before I1 really started expanding, so it's not necessarily a surprise that we find pre-I1 or quasi-I1 lineages in strange places.



I definitely agree that we should stick with caution with how these old samples are labeled indeed. I'm definitely not saying we should ignore the pre-I1 samples, those are likely important, even if they might be extinct lineages, they at least give us a picture of how dispersed I1 related lineages could have been. We know YFull still estimates the formation of I1 to be 27,500 years ago in regards to when it split from I2.

Ah, the old ISOGG 2009 tree, I think that was when FTDNA had finally determined that my bunch weren't just I-M170 and that we were in the I-M253 haplogroup.

Yep, it's frustrating. But as I've said several times before, in most cases Y-DNA analysis isn't the focus of the manuscript - usually just an entry in a table and a couple of sentences in most. I think perhaps the rapid expansion of phylogenetic trees since the introduction of NGS means a lot of folks aren't sure where to look for a reference and fall back to a default which may no longer be applicable.

I'll get through the two Sigtuna samples - just not sure when. These two samples have a lot of no calls so far. I'll try and find as much corroborating information from the BAM files.

You may be surprised if you get round to checking out any BAM files - it's actually a lot easier than I initially expected. The way I'm viewing them is time consuming rather than difficult, but it gives me the best chance of making the right calls.

Hunter Provyn developed Phylogeographer first and yes it's based on the YFull tree rather than the FTDNA tree. I believe that the decision was most likely driven by the fact that at the time of development, the YFull tree was public and access to subclades, age estimates and geographical data was easier for Hunter's data mining programs, while at the time FTDNA's Y-DNA was not public. Since then, Robin Spencer launched SNP Tracker which uses the public FTDNA haplotree as data source for SNPs and subclades, while using the YFull age estimate for the corresponding subclade. There's advantages and disadvantages to both - FTDNA has the larger number of samples but only from it's own customers and also includes folks who have done single SNP tests and SNP packs which may not lead to the terminal SNP on that tree. All of YFull's samples are NGS and so go as far as the terminal can be determined, as well as including other non-FTDNA vendors and academic samples. Both Hunter and Robin are doing good work which is worth checking out.

Agree with the other points.

[spruithean]

Yeah, I tried looking in the "Early Medieval aDNA from Poland Coming Soon" thread on Anthrogenica but over 244 pages comprising 2443 posts, it's difficult to find any reference to actual data buried amongst the conjecture and bickering. Perhaps you could try asking Tomenable in the other forum .

combine that with Anthrogenica's wait time between searches and you'll be hours looking for the right post!

And the haplogroup and subclade assignments for these two samples may very well be correct. But after seeing so many misassignments in publications and secondary analysis of I1 samples - for example: SVK-A1 listed as I1b in the Ebenesersdóttir et al. paper; CL63 listed as I1-Z79 in the Amorim et al paper and numerous holes in the supplementary information regarding calls on that sample and SZ45; Carlos Quiles misassigning HSJ-A1 and SSJ-A2 to I-A8182, SBT-A1 to I-S26062 and repeating the incorrect assignment in the paper of SVK-A1 as I1b, etc.; it's important to check against the raw data for I1 ancient DNA (and indeed all haplogroups).

I mean we're only as good as the data provided by the documents, unless things are double checked. However these corrections should be made at some point so as not to cause confusion down the road.

Yep, it's frustrating. But as I've said several times before, in most cases Y-DNA analysis isn't the focus of the manuscript - usually just an entry in a table and a couple of sentences in most. I think perhaps the rapid expansion of phylogenetic trees since the introduction of NGS means a lot of folks aren't sure where to look for a reference and fall back to a default which may no longer be applicable.

I'll get through the two Sigtuna samples - just not sure when. These two samples have a lot of no calls so far. I'll try and find as much corroborating information from the BAM files.

You may be surprised if you get round to checking out any BAM files - it's actually a lot easier than I initially expected. The way I'm viewing them is time consuming rather than difficult, but it gives me the best chance of making the right calls.

Hunter Provyn developed Phylogeographer first and yes it's based on the YFull tree rather than the FTDNA tree. I believe that the decision was most likely driven by the fact that at the time of development, the YFull tree was public and access to subclades, age estimates and geographical data was easier for Hunter's data mining programs, while at the time FTDNA's Y-DNA was not public. Since then, Robin Spencer launched SNP Tracker which uses the public FTDNA haplotree as data source for SNPs and subclades, while using the YFull age estimate for the corresponding subclade. There's advantages and disadvantages to both - FTDNA has the larger number of samples but only from it's own customers and also includes folks who have done single SNP tests and SNP packs which may not lead to the terminal SNP on that tree. All of YFull's samples are NGS and so go as far as the terminal can be determined, as well as including other non-FTDNA vendors and academic samples. Both Hunter and Robin are doing good work which is worth checking out.

Agree with the other points.

Not only are phylogenetic trees cumbersome now, ISOGG's formatting now has become immensely unruly. YFull certainly has the leg-up on the competition here with the sheer simplicity and the use of the HG-SNP nomenclature. I like the combination of using both websites, they compliment each other I feel, and as time goes on they will improve further with the growth of databases.

[deadly77]

Not only are phylogenetic trees cumbersome now, ISOGG's formatting now has become immensely unruly. YFull certainly has the leg-up on the competition here with the sheer simplicity and the use of the HG-SNP nomenclature. I like the combination of using both websites, they compliment each other I feel, and as time goes on they will improve further with the growth of databases.[/QUOTE]

Yes, that is especially annoying. I don't actually use ISOGG's tree very much any more, and to be honest that is probably more of a reason than they aren't as up to date as YFull or FTDNA. I often load up the page, go and do something else and come back to it several minutes later. And even then I often give up. I find that FTDNA's website and interface is also very clunky and slow which is frustrating, especially trying to navigate the results pages of larger projects - moving to another page of the main I1 project often times out with the "Houston, we have a problem..." error. YFull sometimes slow with loading different branches of the YFull tree although definitely not as slow as FTDNA or ISOGG. Inside the homepage of an account, YFull pretty fast on most things - slowest seems to be while looking up SNPs, positions, STRs in the Groups view.

Rebekah Canada commented on the performance of the YFull website on her Haplogroup blog here: https://haplogroup.org/yfull-launches-new-mtdna-tree/ - she notes that the YFull website scores fairly poorly using Google's Lighthouse tool, but she also notes that despite this, on most computers it loads within a few seconds and also lists the pingdom results.

I wasn't familiar with Google Lighthouse test, but she discusses it here: https://haplogroup.org/genetic-genea...he-lighthouse/ - also serves as a good list of genetic genealogy blogs.

[spruithean]

Never heard of Google Lighthouse. I'll take a look, also thanks for sharing a link to that blog, always interesting to read genetics oriented blogs and websites.

Good to know I'm not the only one struggling to get the I1 DNA Project to load, at least the subclade projects load relatively quickly right now. Hopefully those don't start being problematic.