The scientific studies have a selected list of 12, 17, or 23 markers. However, although due to the presence of Y-STR marker DYS448 (in 17, 23) can be easily predicted I-PH908, the subclades have similar modal haplotypes because of which is difficult to differentiate them without DYS561 and other Y-STR markers. As can be seen in the I-PH908 draft tree (from 23 November 2019), some subclades do have private derivative mutations on Y-STR markers which are characteristic for their prediction, and some of these markers can be found among the selected markers in the studies. Among the selected markers is DYS438 (in 12, 17, 23), which derivation 10 > 9 is seemingly private for I-Y84307 > I-Y93273 (see YFull info).
With that presumption was searched for a PH908 haplotype with an exact mutation on DYS438 which could be a match for the subclade haplotype. It was searched a long list of studies of populations in Central, Eastern, and Southeastern Europe encompassing several thousand samples, including of Albania (Tofanelli 2016), Austria (Berger 2005, Erhart 2012), Belarus (Rębała 2009, Kushniarevich 2013, Pankratov 2016), Bosnia and Herzegovina (Kovačević 2013, Dogan 2016), Bulgaria (Martinez-Cruz 2012, Karachanak 2013), Croatia (Mršić 2012, Šarac 2016, Tofanelli 2016), Czech Republic (Woźniak 2010, Ehler 2011, Purps 2014), Germany (Rodig 2007, Rębała 2013, Purps 2014), Greece (King 2011, Martinez-Cruz 2016, Tofanelli 2016), Hungary (Martinez-Cruz 2012 & 2016, Pamjav 2017), Italy (Purps 2014, Boattini 2015, Tofanelli 2016, Grugni 2018), North Macedonia (Jakovski 2011, Jankova 2019), Moldova (Varzari 2013), Montenegro (Mirabal 2010), Poland (Rębała 2005, Jacewicz 2007-2008, Wolańska-Nowak 2009, Woźniak 2010, Rębała 2013, Purps 2014, Diepenbroek 2019), Romania (Bosch 2006, Stanciu 2010, Bembea 2011, Martinez-Cruz 2012 & 2016), Russia (Roewer 2008, Pesik 2014), Rusyns in Serbia (Rębała 2014, Veselinovic 2014), Serbs in Serbia and near countries (Mirabal 2010, Scorrano 2017, Zgonjanin 2017, Kačar 2019), Slovakia (Petrejčíková 2010, Woźniak 2010, Martinez-Cruz 2016), Slovenia (Zupan 2013, Delser 2018), and Ukraine (Martinez-Cruz 2012 & 2016, Mielnik-Sikorska 2013, Utevska 2015) among others.
A match was found in two samples from Croatia (Mršić 2012), one from Serbia (Mirabal 2010), Hungary (Pamjav 2017), Ukraine (Martinez-Cruz 2012 & 2016), and possibly Romania (Bosch 2006). These six scientific samples were listed with the two samples from YFull currently defining the I-Y84307 > I-Y93273 subclade mainly calculated as one haplotype thus making seven haplotypes in total. In Table 1. are listed haplotypes while in Table 2. are genetic distances between them:
 | |
|
 | |
|
It must be emphasized that Y-STR marker Y-GATA-H4=11 in the AmpFISTR Yfiler PCR Amplification Kit calculation has an equivalent result of Y-GATA-H4=10 in other calculations thus it was edited 11 > 10 in the table because is reducing the initial genetic distance before the tabular inclusion (Table 2.).
Secondly, three haplotypes have or lack some Y-STR markers (Table 1.). While "E818/15" has 23 markers, "UM037" also had DYS426=11 and DYS388=13 which results are shared by YFull haplotype, and "124" also had DYS388=12, DYS436=12, DYS462=12, DYS434=11, DYS461=12, DYS435=11 which results exception of DYS434=9 were shared by YFull haplotype. The latter two additional markers won't be included in the table.
Analysis:
The haplotype of "124" (Romania) has a lack of DYS448=19 to confirm it is I-PH908, but the calculation of Jim Cullen's haplogroup I-subclade predictor gives 40% to be part of Dinaric-South cluster. Additionally, as has a lack of Y-STR markers (13/17) it won't be listed in Table 2., and the specific mutations make its haplotype an isolate, but the closest (4M-5GD) haplotype is of UM037 (Ukraine) with which shares DYS19=17.
According to scientific estimations, the DYS19 has a slow mutation rate, while DYS437 and DYS438 have an even slower rate. This is significant because of the genetic distance between the haplotypes in comparison to "104" (Serbia), ranging from 4M-4GD to 6M-8GD on only 17 Y-STR markers. The sample "104" with DYS437=14, DYS19=16, DYS439=12, DYS458=18, DYS635=23, and even DYS385ab=15-15 does not indicate to be related with other haplotypes, neither as distant branching of the subclade. It is rather a part of a modal haplotype of another currently unknown subclade which is defined by some or most of these marker results (see e.g. SDP). This indicates that the rare STR marker result DYS438=9 is not exclusively present in the I-Y84307 subclade and problematizes the characteristic definition for the same.
As for the genetic distance between the three haplotypes from Croatia, the "71" (Central Croatia) and "72" (Southern Croatia) are very close to the YFull haplotype (1M-2GD, 2M-2GD), and probably could belong to the I-Y84307 subclade. According to YFull and FTDNA analysis of SNPs and STRs, the most recent common ancestor of the two YFull samples is dated to 1250 or 1050, as such the MRCA of these haplotypes lived around or before that time period. The ancestral location was probably located somewhere in a triangle of Western, Central, and Southern regions (Fig. 1). Like the YFull haplotype, the "72" has the same close genetic distance to "E818/15" and "UM037" (2M-2GD, 2M-3GD), while "71" is less close (3M-4GD, 3M-5GD).
Secondly, three haplotypes have or lack some Y-STR markers (Table 1.). While "E818/15" has 23 markers, "UM037" also had DYS426=11 and DYS388=13 which results are shared by YFull haplotype, and "124" also had DYS388=12, DYS436=12, DYS462=12, DYS434=11, DYS461=12, DYS435=11 which results exception of DYS434=9 were shared by YFull haplotype. The latter two additional markers won't be included in the table.
The haplotype of "124" (Romania) has a lack of DYS448=19 to confirm it is I-PH908, but the calculation of Jim Cullen's haplogroup I-subclade predictor gives 40% to be part of Dinaric-South cluster. Additionally, as has a lack of Y-STR markers (13/17) it won't be listed in Table 2., and the specific mutations make its haplotype an isolate, but the closest (4M-5GD) haplotype is of UM037 (Ukraine) with which shares DYS19=17.
According to scientific estimations, the DYS19 has a slow mutation rate, while DYS437 and DYS438 have an even slower rate. This is significant because of the genetic distance between the haplotypes in comparison to "104" (Serbia), ranging from 4M-4GD to 6M-8GD on only 17 Y-STR markers. The sample "104" with DYS437=14, DYS19=16, DYS439=12, DYS458=18, DYS635=23, and even DYS385ab=15-15 does not indicate to be related with other haplotypes, neither as distant branching of the subclade. It is rather a part of a modal haplotype of another currently unknown subclade which is defined by some or most of these marker results (see e.g. SDP). This indicates that the rare STR marker result DYS438=9 is not exclusively present in the I-Y84307 subclade and problematizes the characteristic definition for the same.
As for the genetic distance between the three haplotypes from Croatia, the "71" (Central Croatia) and "72" (Southern Croatia) are very close to the YFull haplotype (1M-2GD, 2M-2GD), and probably could belong to the I-Y84307 subclade. According to YFull and FTDNA analysis of SNPs and STRs, the most recent common ancestor of the two YFull samples is dated to 1250 or 1050, as such the MRCA of these haplotypes lived around or before that time period. The ancestral location was probably located somewhere in a triangle of Western, Central, and Southern regions (Fig. 1). Like the YFull haplotype, the "72" has the same close genetic distance to "E818/15" and "UM037" (2M-2GD, 2M-3GD), while "71" is less close (3M-4GD, 3M-5GD).
 | ||
|
The "E818/15" (Hungary; Bodrogköz, Zemplénagárd) and "UM037" (Ukraine; c. Ivano-Frankivsk and Chernivtsi Oblast) haplotypes are very close to each other (1M-1GD) with an only difference on DYS19. Interestingly enough, the localization of these two haplotypes, with the addition of "124" (Romania; Ploiești), has geographical proximity and gravitation in Eastern Europe (Fig. 2). With the presumption that these three haplotypes are related to the haplotypes from Croatia, the location could suggest the medieval place of origination and movement to the Western Balkans.
 |
| Fig. 2 Map of I-Y84307 with the location (stars) of the three DYS438=9 samples |
Genealogical database samples:
Besides the initially mentioned YFull samples, in genealogical databases exist three more I-PH908 sample haplotypes with DYS438=9. Their haplotypes won't be displayed but will be compared for genetic distance (Table 3.). First is "T1851" from the Czech Republic with 17 markers (see CDP), the second is "101953" from Montenegro with 23 markers and belongs to the I-Y52621 subclade (see SDP), while third is "142396" from Greece with 67 markers (see GDP). As shown, at least two other subclades have one rare sample with a private mutation on DYS438. That problematizes the evaluation of a relationship between haplotypes and whether they belong to the I-Y84307 subclade. However, presuming the relationship and in the context of being in a population located North of Southeastern Europe, on the map is added: "T1851" (Plzeň) which could suggest a different place of origination and movement (Fig. 3).
 | |
|
 |
| Fig. 3 Map of I-Y84307 with the location (stars) of the four DYS438=9 haplotypes |
Y Chromosome Haplotype Reference Database samples:
As most of the scientific haplotypes are in YHRD it will be taken into account 1-2-step neighbor haplotypes of the previously mentioned haplotypes, specifically DYS438=10 with the assumption it is an ancestral mutation of the derivative DYS438=9 and that DYS438=10 haplotypes could also belong to I-Y84307 > I-Y93273 (Table 4.). The haplotypes location, including North of Southeastern Europe, was added on the map (Fig. 4), additionally to them neighbors of "T1851" which is not listed in the haplotype table, while "101953" has closest neighbors from Southeastern Europe. The "101953" and "104" are not placed on the map because of they have private mutations of another subclade. The "142396" lacks the DYS635 marker but nevertheless the result due to private mutations didn't have a match and since it is not known the exact location it was putatively located (blue star). The mentioned "T1851" had 1-step neighbor DYS438=10 from Ukraine (Lviv), Slovenia (Ljubljana) and Croatia (Zagreb), while the 2-step neighbors North of Southeastern Europe in Poland (Warsaw), Germany (Stuttgart) and Russian Federation (Ivanowo).
 | |
|
 | |
|
Conclusion:
In conclusion, the STR marker result DYS438=9 probably is indicative for the PH908 subclade I-Y84307 > I-Y93273, but only if are taken into account results of other STR markers which preceded it like DYS561=15, and are present or absent in the modal haplotype because it is a rare private mutation found in other PH908 subclades. Further STR and SNP sampling are needed to definitively confirm whether the specific marker result is characteristic for all the eight SNPs currently defining the subclade. Among many scientific and genetic genealogical samples were found two very close contemporary haplotypes, both in Croatia, indicating that the nearest genetic matches most probably live in the same country. In short, if it is possible depending on the haplotype of any haplogroup, it is advised to anyone to check both commercial & genealogical and scientific databases & studies for a match as it could give further information on a geographical and historical ancestry.
In conclusion, the STR marker result DYS438=9 probably is indicative for the PH908 subclade I-Y84307 > I-Y93273, but only if are taken into account results of other STR markers which preceded it like DYS561=15, and are present or absent in the modal haplotype because it is a rare private mutation found in other PH908 subclades. Further STR and SNP sampling are needed to definitively confirm whether the specific marker result is characteristic for all the eight SNPs currently defining the subclade. Among many scientific and genetic genealogical samples were found two very close contemporary haplotypes, both in Croatia, indicating that the nearest genetic matches most probably live in the same country. In short, if it is possible depending on the haplotype of any haplogroup, it is advised to anyone to check both commercial & genealogical and scientific databases & studies for a match as it could give further information on a geographical and historical ancestry.
