Replications of posted models, a pre-declared grid of 191 qpAdm models and a robustness check of the Old Hittite (Kaman) fits: a Slavic-type and an Anatolian-type source are needed, the Mycenaean share is not identified.
Mvagnus Albertus · Popualtion Geneticist
Published · 54 min read
Public qpAdm models of Modern Greeks combine a Slavic, a Mycenaean and one Anatolian source and are used to argue about the size of the Aegean share and about which Anatolian proxy is valid. We rebuilt the posted models with proxies from the Nexju base (Human Origins v66 ext 3) and ran a pre-declared grid of 191 models for Modern Greeks (Athens, n=17; Thessaloniki, n=14): the Slavic proxy (Poland Early Medieval, n=46) plus one of 17 Anatolian-type sources, with and without Mycenaeans (n=7), on two right panels. Of the four posted Mycenaean-containing models, only the Old Hittite (Kaman, n=2) model fits, with different weights (Mycenaeans 19 instead of 32 %); the Roman-era, Alalakh and Arslantepe models are rejected. No model without a Slavic source (0 of 34) and no model without an Anatolian source (0 of 4) fits. Without Mycenaeans, 10 two-source models fit; 6 of them become infeasible when Mycenaeans are added. Only 2 of 68 three-source models are strict fits, and both use the same source of two persons (Old Hittite, Kaman). A second pre-declared block (116 models and 4 sensitivity analyses) tested these two fits. Removing one of the two persons leaves a rejected model (p = 0.040 for Athens, 0.029 for Thessaloniki), raw instead of reduced counts of two right populations reject the Athens fit (p = 0.037), and the fit depends on the right panel. By the pre-declared rule the Kaman three-source fits are INVALID: they are not used as evidence about Greeks. This does not show them to be false. The Mycenaean share is not identified: passing models with a Kaman-site source give 5.3 to 50.2 %.
qpAdm1 tests whether the allele frequencies of a target population can be reproduced as a mixture of a set of source populations, relative to a set of right (outgroup) populations. A model is not rejected when the p-value of the fit is not below 0.05. That is the absence of rejection with the power of the data, not evidence of descent.2
A public discussion about modern Thessalonians and Athenians uses qpAdm models with three sources: a Slavic-type source (Early Medieval Central and Eastern Europe), a Mycenaean or Aegean Bronze Age source, and one Anatolian-type source. The posted models differ in the Anatolian source (Roman-era, Alalakh, Arslantepe, Old Hittite). The arguments around them are (i) that the Aegean share of Greeks is large (“well over 35 %”), (ii) that Alalakh and Arslantepe are valid proxies of post-Chalcolithic Anatolian ancestry whereas Old Hittite or Early Bronze Age Anatolians do not fit, and (iii) that, when closer West-Anatolian Bronze Age sources are used, the drift goes to them and not to the Mycenaean source.
This study asks four questions. (Q1) Can the posted models be reproduced with the sources of the Nexju base? (Q2) How do fit, weights and the Mycenaean share react when the Anatolian source is changed, and does a model need a Slavic, an Anatolian and a Mycenaean source? (Q3) What is the evidence for the Old Hittite (Kaman) models, the only strict three-source fits? (Q4) Does that evidence survive checks that were fixed before the runs? The rules for all comparisons were written down before the runs of each block and were not changed afterwards.
Data and targets. All models use the Nexju base (Human Origins v66 ext 3, with individual labels that follow the Allen Ancient DNA Resource3), qpAdm with allSnps false unless stated, α = 0.05. The targets are the Modern Greek groups Greek_WGA (Athens, n=17) and Greek_Thessaloniki (n=14). The Slavic proxy (SP) is Poland Early Medieval Slavs (n=46); the Mycenaean source (MY) is a group of seven Mycenaean individuals (n=7). The 17 Anatolian-type sources X are nine Roman-era groups, two Byzantine and one Ottoman group, Alalakh (n=5), Arslantepe (n=16), Old Hittite from Kalehöyük, Kaman (n=2; KAM2: MA2200, 1,021,939 SNPs, and MA2203, 317,679 SNPs), and two West-Anatolian Bronze Age groups. Table 14 lists every group with its code and members. Several of these groups are small (n=1 to 5) or have fewer than 500,000 SNPs; this is stated where it matters.
Right panels. Two panels are used. RB13 (13 rights) is built from the base populations that correspond to the 13 outgroup classes of the posted models; R10 (10 rights) is the panel of earlier Nexju blocks. Both are proxies of the posted panels, not identical ones. R14 is R10 plus four populations of the Anatolian and Levantine side. Two right populations hold fewer individuals in the datasets of block 1 than their base label suggests (Turkey_N 43 instead of 68 on RB13, Crete 5 instead of 34); this is tested in the check V6.
Fit rule. A model is a fit if p ≥ 0.05, every weight lies in [0,1], and every weight standard error is below 0.10. A model with p ≥ 0.05 and feasible weights but a standard error of 0.10 or more is imprecise; with a weight outside [0,1] it is infeasible. A published-style pass needs only p ≥ 0.05 and weights in [0,1]. The fits to the posted models M1–M4 use allSnps true, as in the posts. allSnps true stretches the models artificially and distorts p-values, so the p-values of M1–M4 are not comparable with the allSnps false results of the other blocks.
Record check. Every model is a single qpAdm run with exactly the sources of the model on the left, stored with its analysis id; every record was checked (sources and right count as declared, degrees of freedom = rights − sources, weights sum to 1, p reproduced from χ² and degrees of freedom). The tables are generated from the records. Each model in the tables and appendices carries a link to its public Nexju result page.
Two blocks were run. Block 1 (191 single models): replications of the posted models M1–M4 (Papadopoulos) and CA1–CA3 (Antoniou), the grid SP + X (+ MY) for the 17 sources X on both panels and both targets (136 models), models without a Slavic source (34), without an Anatolian source (4), and a test of the replications on raw counts (7). Block 2 (116 single models and 4 sensitivity analyses): the checks of the Kaman fits (Table 1). The rules, the groups, the control pairs (drawn with a fixed random seed) and the expectations were written down before the runs of each block; counts and verdicts below are computed by those rules. Nothing was re-ranked afterwards, and nothing was corrected for multiplicity.
The verdict rule is: INVALID if V1 is fragile for at least one target, or V2 is fragile for at least one target, or V6 does not hold, or (V4 is “none” and V5 is not discriminating for at least one target). Otherwise valid for Kaman only if V4 is “none”, or V5 is weak or not discriminating, or V2 is mixed for at least one target. Otherwise robust. “Invalid” means that the fit is not used as evidence for any statement about Greeks; it stays in the record as a result of this source and this panel.
Table 1 | Checks V1–V6 fixed before the Kaman block
Check | Question | Rule (per target) |
|---|---|---|
V1 | Is the fit carried by one person? | Slavs + MA2200 + MY and Slavs + MA2203 + MY (RB13) must both be published-style passes; otherwise fragile. |
V2 | Does it survive other right sets? | 14 variants of Slavs + KAM2 + MY (12 drop-one-right, panels R10 and R14): stable if at least 12 are not rejected, mixed if 10 or 11, fragile if 9 or fewer. |
V3 | Is the Mycenaean source needed? | Slavs + KAM2 rejected in at least 12 of its 14 variants and the drop-source variant rejected: required. |
V4 | Does an independent source fit too? | Strict three-source fit on RB13 with Kaman Assyrian Colony (n=3), Kaman Iron Age (n=2), Oylum Höyük (n=6) or Harmanoğlu (n=3): corroborated. |
V5 | Is a pass specific to this pair? | Eight control pairs of persons from pools rejected as whole groups: 0 or 1 pass = discriminating, 2 = weak, 3 or more = not discriminating. |
V6 | Does it depend on two right counts? | Slavs + KAM2 + MY with raw counts (Turkey_N 68, Crete 34) is a published-style pass for both targets: holds. |
Table 2 compares the posted models with the Nexju versions (target Thessaloniki, RB13). The Slavic weight is reproduced in M2–M4 (within 2.3 points), but the split between Mycenaean and Anatolian weight is not. M1 (Roman-era plus Ottoman, four sources) is rejected and infeasible (the two Anatolian weights are 123 and −118 %). M2 (Alalakh) and M3 (Arslantepe) are rejected (p = 2 × 10⁻⁵ and 2 × 10⁻¹¹), although they are the models presented as fitting. M4 (Old Hittite, Kaman) fits (p = 0.071 against the posted 0.164) with Mycenaeans at 19 % instead of 32 %. The models of Antoniou (CA1, CA2) pass, but they rest on 27,781 and 12,701 SNPs with one Aydın individual (82,389 SNPs) and the standard errors exceed the weights; a source this poor fits almost anything, so these passes carry little information. The replication is not exact: tool, SNP set, individuals and right populations differ from the posts. The fit of M1–M4 does not change when the raw counts of Turkey_N and Crete are used (largest weight change 3.6 points; models RZ in Appendix D).
Table 2 | Posted models and their replication with Nexju proxies, target Thessaloniki (n=14), RB13
Model | Target | p posted | p ours | Result | Largest weight difference (points) | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|---|
M1 Roman-era + Ottoman | Thessaloniki | 0.427 | 4 × 10⁻⁵ † | rejected (infeasible) | n/a (infeasible) | 468,442 | MY 73.0 (20.5), Q10 123.4 (58.3), SP 21.2 (9.8), OTT5 −117.6 (67.1) | |
M2 Alalakh | Thessaloniki | 0.384 | 2 × 10⁻⁵ † | rejected | 11.6 | 468,442 | MY 26.5 (5.0), ALA5 28.4 (4.3), SP 45.0 (1.5) | |
M3 Arslantepe | Thessaloniki | 0.389 | 2 × 10⁻¹¹ † | rejected | 1.8 | 468,442 | MY 33.8 (6.3), ARS16 21.1 (5.1), SP 45.2 (1.8) | |
M4 Old Hittite (Kaman) | Thessaloniki | 0.164 | 0.071 † | fit | 14.9 | 468,442 | MY 19.4 (6.0), KAM2 39.0 (5.8), SP 41.6 (1.8) | |
CA1 Lazarides LBA + Aydın BA | Thessaloniki | 0.448 | 0.653 | pass, imprecise | 6.7 | 27,781 | LAZ4 11.4 (480.9), LEB 4.6 (68.2), CZ 34.7 (82.9), AYD1 49.3 (630.2) | |
CA2 Lazarides EBA + Aydın BA | Thessaloniki | 0.461 | 0.959 | pass, imprecise | 31.6 | 12,701 | LAZ1 13.8 (8.9), LEB 28.5 (16.6), CZ 43.0 (8.9), AYD1 14.7 (26.6) |
† allSnps true, as in the posts. It stretches the models and distorts p-values, so these p-values are not comparable with allSnps false results. The CA models use allSnps false. “Largest weight difference” is in percentage points against the posted figure. Codes as in Table 14.
Table 3 summarizes the 191 models of block 1 by class. Three results hold in all cells that were run.
A Slavic-type source is needed. None of the 34 models X + Mycenaeans fits; 28 of the 34 are infeasible.
An Anatolian-type source is needed. Slavs + Mycenaeans is rejected on all four cells (p at most 3 × 10⁻¹¹).
Strict three-source fits are rare. Of 68 models Slavs + X + Mycenaeans only 2 are fits, both with X = Kaman (RB13). The same sources without Mycenaeans give 10 fits.
Table 3 | Number of models of block 1 by class and result
Class | Models | Fit | Imprecise | Infeasible, p ≥ 0.05 | Rejected |
|---|---|---|---|---|---|
Slavs + X + Mycenaeans (17 X, 2 targets, 2 panels) | 68 | 2 | 4 | 6 | 56 |
Slavs + X (no Mycenaeans) | 68 | 10 | 1 | 0 | 57 |
X + Mycenaeans (no Slavic source) | 34 | 0 | 0 | 2 | 32 |
Slavs + Mycenaeans (no Anatolian source) | 4 | 0 | 0 | 0 | 4 |
Fit = strict rule; imprecise and infeasible = p ≥ 0.05 but not a fit; rejected includes rejected and infeasible.
Fits without a Mycenaean source. Table 4 lists the 10 strict two-source fits. All involve small groups: Aydın Bronze Age (n=1), Turkey Imperial Roman (n=4), Roman-era QW2 (n=2) and QPW5 (n=5, assembled after the first results, hence post hoc), and Roman-era QP3 (n=3). Adding Mycenaeans raises p in 56 of 68 pairs (a free parameter is added) but turns the strict fits into: 6 infeasible (negative Mycenaean weight −1 to −22 %), 1 rejected (QP3, Athens, Mycenaean 0.7 %) and 3 imprecise (Aydın, n=1). The Slavic weight is stable in the models that are not rejected: 27 to 34 % for Athens and 37 to 42 % for Thessaloniki (Aydın models left out).
Table 4 | The strict fits without a Mycenaean source (Slavic proxy + one Anatolian group)
Target | Anatolian source (with Slavs) | Panel | p | Weights % (SE) | SNPs | Link |
|---|---|---|---|---|---|---|
Thessaloniki | Aydın Bronze Age (n=1) | R10 | 0.448 | SP 35.9 (10.0), AYD1 64.1 (10.0) | 31,719 | |
Thessaloniki | Turkey Imperial Roman (n=4) | R10 | 0.522 | SP 36.7 (2.9), O4 63.3 (2.9) | 333,884 | |
Thessaloniki | Turkey Roman-era QPW5 (n=5) | R10 | 0.060 | SP 40.6 (2.2), QPW5 59.4 (2.2) | 424,172 | |
Thessaloniki | Turkey Roman-era QW2 (n=2) | R10 | 0.068 | SP 41.3 (2.6), QW2 58.7 (2.6) | 407,319 | |
Thessaloniki | Aydın Bronze Age (n=1) | RB13 | 0.864 | SP 26.1 (6.1), AYD1 73.9 (6.1) | 30,599 | |
Athens | Turkey Imperial Roman (n=4) | R10 | 0.254 | SP 27.0 (3.1), O4 73.0 (3.1) | 333,884 | |
Athens | Turkey Roman-era QPW5 (n=5) | R10 | 0.138 | SP 31.4 (2.2), QPW5 68.6 (2.2) | 424,172 | |
Athens | Turkey Roman-era QW2 (n=2) | R10 | 0.199 | SP 34.0 (2.6), QW2 66.0 (2.6) | 407,319 | |
Athens | Aydın Bronze Age (n=1) | RB13 | 0.809 | SP 17.8 (6.9), AYD1 82.2 (6.9) | 30,599 | |
Athens | Turkey Roman-era QP3 (n=3) | RB13 | 0.057 | SP 32.4 (1.9), QP3 67.6 (1.9) | 395,400 |
allSnps false. Codes as in Table 14.
Sources of the grid. The p-values of the whole grid are in Tables 5 and 6; every model with weights and link is in Appendix D. Roman-era Anatolian groups give no three-source fit: they are rejected or, in six cells, feasible only with a negative Mycenaean weight. Alalakh and Arslantepe are rejected in the three-source models on both targets, with p of 9 × 10⁻⁶ and 2 × 10⁻¹³ (Athens) and 4 × 10⁻⁵ and 3 × 10⁻⁹ (Thessaloniki). The Mycenaean weight of the 52 feasible three-source models runs from 0.7 to 65 % depending on the Anatolian source. Of the 12 three-source models with p ≥ 0.05, six are infeasible, four imprecise and two are fits.
Table 5 | Grid p-values, Modern Greeks (Athens, n=17)
Source X | RB13 Slavs + X | RB13 + MY | R10 Slavs + X | R10 + MY |
|---|---|---|---|---|
Turkey Roman-era (n=10) | <1e-16 R | 7e-11 R | 2e-14 R | 3e-6 R |
Turkey Roman-era QI5 (n=5) | 7e-8 R | 3e-5 R | 6e-5 R | 3e-4 R |
Turkey Roman-era QL5 (n=5) | <1e-16 RX | 1e-12 R | <1e-16 RX | 3e-6 R |
Turkey Roman-era QP3 (n=3) | 0.057 F | 0.031 R | 0.045 R | 0.039 RX |
Turkey Roman-era QB3 (n=3) | <1e-16 RX | 7e-16 R | <1e-16 RX | 1e-5 RX |
Turkey Roman-era QW2 (n=2) | 5e-9 R | 9e-6 R | 0.199 F | 0.265 X |
Turkey Roman-era QS2 (n=2) | 1e-16 R | 4e-6 R | 1e-7 R | 8e-5 R |
Turkey Roman-era QPW5 (n=5) | 0.018 R | 0.024 R | 0.138 F | 0.217 X |
Turkey Imperial Roman (n=4) | 0.006 R | 0.008 RX | 0.254 F | 0.396 X |
Byzantine West Anatolia early (n=21) | 2e-12 R | 3e-12 RX | 1e-4 R | 2e-4 RX |
Byzantine West Anatolia late (n=8) | 1e-13 R | 3e-14 R | 6e-6 R | 9e-6 RX |
Turkey Ottoman (n=5) | <1e-16 R | 1e-13 R | 5e-8 R | 8e-7 R |
Alalakh MLBA (n=5) | 3e-10 R | 9e-6 R | 2e-8 R | 8e-7 R |
Arslantepe LC–EBA (n=16) | <1e-16 R | 2e-13 R | 8e-12 R | 5e-11 R |
Kaman (Old Hittite, n=2) | 0.018 R | 0.128 F | 0.045 R | 0.047 R |
Yassitepe BA (n=2) | 2e-9 R | 0.004 RX | 4e-9 R | 0.003 RX |
Aydın Bronze Age (n=1) | 0.809 F | 0.714 I | 0.625 I | 0.547 I |
Each cell: p and result (F fit, I imprecise, X infeasible with p ≥ 0.05, R rejected, RX rejected and infeasible). Per panel, the pair is Slavs + X without and with Mycenaeans (+ MY). allSnps false.
Table 6 | Grid p-values, Modern Greeks (Thessaloniki, n=14)
Source X | RB13 Slavs + X | RB13 + MY | R10 Slavs + X | R10 + MY |
|---|---|---|---|---|
Turkey Roman-era (n=10) | <1e-16 R | 4e-8 R | <1e-16 R | 1e-5 R |
Turkey Roman-era QI5 (n=5) | 3e-8 R | 1e-4 R | 4e-7 R | 7e-5 R |
Turkey Roman-era QL5 (n=5) | <1e-16 R | 2e-9 R | <1e-16 RX | 1e-5 R |
Turkey Roman-era QP3 (n=3) | 0.036 R | 0.023 R | 0.019 R | 0.027 R |
Turkey Roman-era QB3 (n=3) | <1e-16 RX | 5e-11 R | <1e-16 RX | 2e-5 R |
Turkey Roman-era QW2 (n=2) | 5e-10 R | 9e-5 R | 0.068 F | 0.059 X |
Turkey Roman-era QS2 (n=2) | <1e-16 R | 8e-6 R | 5e-11 R | 5e-6 R |
Turkey Roman-era QPW5 (n=5) | 6e-4 R | 0.005 R | 0.060 F | 0.065 X |
Turkey Imperial Roman (n=4) | 0.005 R | 0.004 R | 0.522 F | 0.497 X |
Byzantine West Anatolia early (n=21) | 3e-11 R | 1e-9 R | 4e-6 R | 3e-6 R |
Byzantine West Anatolia late (n=8) | 3e-13 R | 9e-10 R | 1e-7 R | 2e-7 R |
Turkey Ottoman (n=5) | <1e-16 R | 7e-9 R | 1e-9 R | 8e-6 R |
Alalakh MLBA (n=5) | 9e-12 R | 4e-5 R | 2e-11 R | 5e-6 R |
Arslantepe LC–EBA (n=16) | <1e-16 R | 3e-9 R | 3e-13 R | 4e-9 R |
Kaman (Old Hittite, n=2) | 0.006 R | 0.083 F | 0.002 R | 0.003 R |
Yassitepe BA (n=2) | 1e-7 R | 0.010 RX | 3e-8 R | 0.001 RX |
Aydın Bronze Age (n=1) | 0.864 F | 0.859 I | 0.448 F | 0.339 I |
Cells as in Table 5. allSnps false.
The Kaman three-source models are declared INVALID by the pre-declared rule of this study (see the verdict below). They are shown here because they are the only strict three-source fits; they are not used as evidence for any statement about Greeks.
The strict fits are Slavs + Kaman + Mycenaeans on RB13: Athens p = 0.128 (Slavs 31.3 %, Kaman 47.2 %, Mycenaeans 21.5 %, SE 6.7) and Thessaloniki p = 0.083 (Slavs 41.8 %, Kaman 35.3 %, Mycenaeans 22.8 %, SE 6.5). Table 7 shows the neighbouring models. On R10 the same model is rejected (Athens p = 0.047, a hair under the line; Thessaloniki p = 0.003); without Mycenaeans it is rejected on both panels (RB13: 0.018 and 0.006; R10: 0.045 and 0.002). The Kaman group is two persons, one of them with 318,000 SNPs.
Table 7 | Kaman (Old Hittite, n=2) three-source models and their neighbours
Target | Model | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
Athens | Slavs + Kaman + Mycenaeans | RB13 | 0.128 | fit | 369,507 | SP 31.3 (1.9), KAM2 47.2 (6.3), MY 21.5 (6.7) | |
Athens | Slavs + Kaman + Mycenaeans | R10 | 0.047 | rejected | 387,656 | SP 32.7 (2.7), KAM2 55.3 (8.1), MY 12.0 (8.1) | |
Athens | Slavs + Kaman | RB13 | 0.018 | rejected | 373,260 | SP 33.0 (2.4), KAM2 67.0 (2.4) | |
Athens | Slavs + Kaman | R10 | 0.045 | rejected | 389,918 | SP 33.2 (3.1), KAM2 66.8 (3.1) | |
Athens | Slavs + Kaman + Mycenaeans, raw counts of Turkey_N and Crete | RB13 | 0.037 | rejected | 371,818 | SP 32.7 (2.0), KAM2 49.9 (6.8), MY 17.5 (7.2) | |
Thessaloniki | Slavs + Kaman + Mycenaeans | RB13 | 0.083 | fit | 369,507 | SP 41.8 (1.8), KAM2 35.3 (6.2), MY 22.8 (6.5) | |
Thessaloniki | Slavs + Kaman + Mycenaeans | R10 | 0.003 | rejected | 387,656 | SP 40.8 (2.6), KAM2 45.0 (8.7), MY 14.2 (8.5) | |
Thessaloniki | Slavs + Kaman | RB13 | 0.006 | rejected | 373,260 | SP 43.4 (2.3), KAM2 56.6 (2.3) | |
Thessaloniki | Slavs + Kaman | R10 | 0.002 | rejected | 389,918 | SP 41.2 (3.0), KAM2 58.8 (3.0) | |
Thessaloniki | Slavs + Kaman + Mycenaeans, raw counts of Turkey_N and Crete | RB13 | 0.155 | fit | 371,818 | SP 42.5 (1.8), KAM2 34.3 (6.1), MY 23.2 (6.3) |
SP Slavs, KAM2 Kaman (n=2), MY Mycenaeans. allSnps false.
V1: one person at a time (Table 8). Without MA2200 (the person with 1,021,939 SNPs) the model Slavs + MA2203 + MY passes (Athens p = 0.348, Thessaloniki p = 0.314). Without MA2203 it is rejected: Slavs + MA2200 + MY gives p = 0.040 for Athens and 0.029 for Thessaloniki, with Mycenaeans at about 20 %. MA2203 alone has only 114,164 SNPs on the panel (in the three-source model), and Slavs + MA2203 passes even without Mycenaeans (p 0.128 and 0.062). The pair passes more easily than its better-covered member (speculation, not tested: the poorly covered person widens the standard errors of the group frequencies). V1 is fragile for both targets.
Table 8 | V1: the Kaman model with both persons, with MA2200 only and with MA2203 only (RB13)
Target | Model | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
Athens | Slavs + Kaman (both persons) + Mycenaeans | RB13 | 0.128 | fit | 369,507 | SP 31.3 (1.9), KAM2 47.2 (6.3), MY 21.5 (6.7) | |
Athens | Slavs + MA2200 + Mycenaeans | RB13 | 0.040 | rejected | 359,242 | SP 32.1 (2.2), K00 47.7 (7.6), MY 20.2 (8.1) | |
Athens | Slavs + MA2203 + Mycenaeans | RB13 | 0.348 | fit | 114,164 | SP 30.2 (2.9), K03 46.8 (8.3), MY 23.0 (8.6) | |
Athens | Slavs + MA2200 | RB13 | 0.014 | rejected | 362,904 | SP 33.7 (2.6), K00 66.3 (2.6) | |
Athens | Slavs + MA2203 | RB13 | 0.128 | fit | 115,179 | SP 32.1 (3.8), K03 67.9 (3.8) | |
Thessaloniki | Slavs + Kaman (both persons) + Mycenaeans | RB13 | 0.083 | fit | 369,507 | SP 41.8 (1.8), KAM2 35.3 (6.2), MY 22.8 (6.5) | |
Thessaloniki | Slavs + MA2200 + Mycenaeans | RB13 | 0.029 | rejected | 359,242 | SP 42.2 (2.0), K00 35.9 (7.6), MY 21.9 (8.0) | |
Thessaloniki | Slavs + MA2203 + Mycenaeans | RB13 | 0.314 | fit | 114,164 | SP 39.7 (2.7), K03 34.3 (8.5), MY 26.0 (8.7) | |
Thessaloniki | Slavs + MA2200 | RB13 | 0.007 | rejected | 362,904 | SP 43.9 (2.5), K00 56.1 (2.5) | |
Thessaloniki | Slavs + MA2203 | RB13 | 0.062 | fit | 115,179 | SP 41.6 (3.6), K03 58.4 (3.6) |
K00 Kaman person MA2200 (1,021,939 SNPs), K03 Kaman person MA2203 (317,679 SNPs). allSnps false.
V2: right populations (Table 9; every variant in Appendix C). Athens: the base fit is not rejected in 11 of 14 variants (mixed); it is rejected when Ukraine_EIA_Scythian is dropped (p = 0.034), under R10 (p = 0.047) and under R14 (p = 0.001). Thessaloniki: not rejected in 12 of 14 (stable), rejected under R10 (p = 0.003) and R14 (p = 2 × 10⁻⁴). Inside RB13 the Mycenaean split is stable (15.9 to 23.1 % in the 11 Athens variants and 18.6 to 25.0 % in the 12 Thessaloniki variants); the change comes with the other panels.
V3: Mycenaean source needed in the Kaman model. Slavs + Kaman without Mycenaeans is rejected in 12 of 14 right-panel variants for Athens and in 14 of 14 for Thessaloniki; the drop-source variant is rejected (p = 0.018 and 0.006). Inside this model the Mycenaean source is required. This does not carry over to other sources (Tables 5 and 6).
Table 9 | V2 and V3: the Kaman model against the right populations (14 variants per model)
Target | Model | Variants not rejected | Rejected variants (p) | Mycenaean % in the not-rejected variants | Drop-source variants (p) |
|---|---|---|---|---|---|
Athens | Slavs + Kaman + Mycenaeans | 11 of 14 (mixed) | Ukraine_EIA_Scythian dropped (0.034), R10 (0.047), R14 (0.001) | 15.9 to 23.1 | no Slavs 1 × 10⁻¹⁵ (infeasible); no Kaman < 1 × 10⁻¹⁶; no Mycenaeans 0.018 |
Thessaloniki | Slavs + Kaman + Mycenaeans | 12 of 14 (stable) | R10 (0.003), R14 (2 × 10⁻⁴) | 18.6 to 25.0 | no Slavs 1 × 10⁻¹⁶ (infeasible); no Kaman 3 × 10⁻¹¹; no Mycenaeans 0.006 |
Athens | Slavs + Kaman | 2 of 14 (Croatia_MLBA dropped 0.097, Estonia_LBA_StoneCistGraves dropped 0.125) | the other 12 | – | no Slavs < 1 × 10⁻¹⁶; no Kaman < 1 × 10⁻¹⁶ |
Thessaloniki | Slavs + Kaman | 0 of 14 | all 14 | – | no Slavs < 1 × 10⁻¹⁶; no Kaman < 1 × 10⁻¹⁶ |
The 14 variants are the 12 drop-one-right variants of RB13 and the panels R10 and R14. allSnps false.
V4: other Kaman-site and Anatolian groups (Table 10). Kaman Iron Age (n=2; 194,713 SNPs) gives a strict fit for both targets (Athens p = 0.173, Thessaloniki p = 0.146), so V4 is corroborated. The composition differs strongly from the Old Hittite fit: Mycenaeans 50.2 and 46.9 % (SE 3.6 and 3.1) against about 22 %, and the Slavic weight 17 and 33 %. So a second Kaman-site group can also produce a fit with a Mycenaean source, but not the same weights. Kaman Iron Age is rejected for Athens on R10 (p = 0.033). Kaman Assyrian Colony (n=3) is rejected on RB13 (p = 0.014 and 0.037) and fits on R10 with Mycenaeans at 5.3 and 11.2 %. Oylum Höyük (n=6) and Harmanoğlu (n=3) are rejected on both panels (Oylum p = 1 × 10⁻⁵ down to 1 × 10⁻⁸). All seven Kaman individuals together (not independent of KAM2) fit for Athens (RB13 p = 0.052) and are rejected for Thessaloniki (p = 0.024).
Table 10 | V4: Slavs + other Kaman-site group + Mycenaeans, both panels
Target | Model | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
Athens | Slavs + Kaman Iron Age (n=2) + Mycenaeans | RB13 | 0.173 | fit | 194,713 | SP 17.2 (2.8), KIA2 32.6 (5.1), MY 50.2 (3.6) | |
Athens | Slavs + Kaman Iron Age (n=2) + Mycenaeans | R10 | 0.033 | rejected | 203,370 | SP 12.2 (4.1), KIA2 38.7 (5.0), MY 49.1 (3.4) | |
Thessaloniki | Slavs + Kaman Iron Age (n=2) + Mycenaeans | RB13 | 0.146 | fit | 194,713 | SP 33.2 (2.8), KIA2 19.9 (4.7), MY 46.9 (3.1) | |
Thessaloniki | Slavs + Kaman Iron Age (n=2) + Mycenaeans | R10 | 0.084 | fit | 203,370 | SP 27.8 (3.9), KIA2 28.4 (4.8), MY 43.8 (3.0) | |
Athens | Slavs + Kaman Assyrian Colony (n=3) + Mycenaeans | RB13 | 0.014 | rejected | 303,961 | SP 32.7 (1.8), AC3 42.4 (5.8), MY 24.9 (6.5) | |
Athens | Slavs + Kaman Assyrian Colony (n=3) + Mycenaeans | R10 | 0.122 | fit | 318,702 | SP 37.2 (2.6), AC3 57.5 (6.8), MY 5.3 (7.5) | |
Thessaloniki | Slavs + Kaman Assyrian Colony (n=3) + Mycenaeans | RB13 | 0.037 | rejected | 303,961 | SP 43.2 (1.8), AC3 32.7 (5.5), MY 24.0 (6.1) | |
Thessaloniki | Slavs + Kaman Assyrian Colony (n=3) + Mycenaeans | R10 | 0.065 | fit | 318,702 | SP 45.0 (2.4), AC3 43.8 (6.0), MY 11.2 (6.5) | |
Athens | Slavs + Kaman all periods (n=7) + Mycenaeans | RB13 | 0.052 | fit | 396,446 | SP 28.6 (1.4), KAL7 41.7 (4.4), MY 29.7 (4.4) | |
Athens | Slavs + Kaman all periods (n=7) + Mycenaeans | R10 | 0.064 | fit | 416,864 | SP 29.7 (2.0), KAL7 53.0 (4.9), MY 17.2 (4.7) | |
Thessaloniki | Slavs + Kaman all periods (n=7) + Mycenaeans | RB13 | 0.024 | rejected | 396,446 | SP 40.0 (1.5), KAL7 31.4 (4.4), MY 28.6 (4.6) | |
Thessaloniki | Slavs + Kaman all periods (n=7) + Mycenaeans | R10 | 0.006 | rejected | 416,864 | SP 39.3 (2.0), KAL7 39.6 (4.8), MY 21.1 (4.6) | |
Athens | Slavs + Oylum Höyük MBA (n=6) + Mycenaeans | RB13 | 1 × 10⁻⁵ | rejected | 398,364 | SP 34.9 (1.5), OYL6 32.6 (4.1), MY 32.5 (5.0) | |
Athens | Slavs + Oylum Höyük MBA (n=6) + Mycenaeans | R10 | 1 × 10⁻⁸ | rejected | 419,646 | SP 37.1 (1.8), OYL6 39.8 (4.3), MY 23.1 (4.9) | |
Thessaloniki | Slavs + Oylum Höyük MBA (n=6) + Mycenaeans | RB13 | 8 × 10⁻⁵ | rejected | 398,364 | SP 44.2 (1.6), OYL6 22.6 (3.8), MY 33.2 (4.6) | |
Thessaloniki | Slavs + Oylum Höyük MBA (n=6) + Mycenaeans | R10 | 1 × 10⁻⁶ | rejected | 419,646 | SP 44.3 (1.8), OYL6 26.9 (3.6), MY 28.8 (4.1) | |
Athens | Slavs + Harmanoğlu–Göndürle EBA (n=3) + Mycenaeans | RB13 | 4 × 10⁻⁹ | rejected (infeasible) | 395,104 | SP 61.0 (25.0), HAR3 238.9 (201.7), MY −199.9 (225.8) | |
Athens | Slavs + Harmanoğlu–Göndürle EBA (n=3) + Mycenaeans | R10 | 3 × 10⁻⁷ | rejected (infeasible) | 418,542 | SP 40.2 (3.8), HAR3 94.4 (21.2), MY −34.6 (23.2) | |
Thessaloniki | Slavs + Harmanoğlu–Göndürle EBA (n=3) + Mycenaeans | RB13 | 2 × 10⁻⁹ | rejected (infeasible) | 395,104 | SP −0.7 (259.4), HAR3 −359.3 (2172.3), MY 460.0 (2431.3) | |
Thessaloniki | Slavs + Harmanoğlu–Göndürle EBA (n=3) + Mycenaeans | R10 | 7 × 10⁻⁹ | rejected (infeasible) | 418,542 | SP 46.2 (3.4), HAR3 65.9 (26.3), MY −12.1 (28.3) |
allSnps false. Codes as in Table 14.
V5: do generic pairs pass? (Table 11). None of the eight control pairs (persons of Roman-era Anatolia and Arslantepe, groups rejected as wholes) passes Slavs + pair + Mycenaeans on RB13 for either target (p from 1 × 10⁻¹⁵ to 0.020). A pass is therefore not a generic property of two-person groups on this panel. The controls come from pools that were already rejected as whole groups, which makes them a conservative control in the direction of “discriminating”. On R10 without Mycenaeans, 2 of 8 pairs pass for Athens and 1 of 8 for Thessaloniki, so small groups can pass the two-source model on R10.
Table 11 | V5: eight control pairs (n=2 each), Slavs + pair + Mycenaeans on RB13 and Slavs + pair on R10
Target | Model | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
Athens | Slavs + CQ1 + Mycenaeans | RB13 | 8 × 10⁻⁵ | rejected | 373,765 | SP 31.5 (1.7), CQ1 45.5 (8.8), MY 23.1 (9.1) | |
Thessaloniki | Slavs + CQ1 + Mycenaeans | RB13 | 8 × 10⁻⁵ | rejected | 373,765 | SP 41.4 (1.5), CQ1 29.9 (8.1), MY 28.7 (8.5) | |
Athens | Slavs + CQ2 + Mycenaeans | RB13 | 0.020 | rejected | 380,744 | SP 32.8 (2.3), CQ2 62.8 (12.0), MY 4.4 (12.7) | |
Thessaloniki | Slavs + CQ2 + Mycenaeans | RB13 | 0.008 | rejected | 380,744 | SP 43.0 (2.1), CQ2 45.1 (11.0), MY 11.9 (11.6) | |
Athens | Slavs + CQ3 + Mycenaeans | RB13 | 9 × 10⁻⁶ | rejected | 379,530 | SP 34.9 (1.7), CQ3 34.7 (6.6), MY 30.4 (7.5) | |
Thessaloniki | Slavs + CQ3 + Mycenaeans | RB13 | 9 × 10⁻⁵ | rejected | 379,530 | SP 43.3 (1.6), CQ3 23.4 (5.6), MY 33.3 (6.3) | |
Athens | Slavs + CQ4 + Mycenaeans | RB13 | 1 × 10⁻¹⁵ | rejected | 401,630 | SP 21.7 (3.5), CQ4 9.8 (3.2), MY 68.5 (1.7) | |
Thessaloniki | Slavs + CQ4 + Mycenaeans | RB13 | 6 × 10⁻¹¹ | rejected | 401,630 | SP 37.1 (3.4), CQ4 4.7 (3.1), MY 58.2 (1.7) | |
Athens | Slavs + CR1 + Mycenaeans | RB13 | 8 × 10⁻⁵ | rejected (infeasible) | 335,134 | SP 42.7 (3.6), CR1 65.8 (16.7), MY −8.5 (19.7) | |
Thessaloniki | Slavs + CR1 + Mycenaeans | RB13 | 2 × 10⁻⁴ | rejected | 335,134 | SP 49.7 (2.9), CR1 44.7 (12.4), MY 5.6 (14.7) | |
Athens | Slavs + CR2 + Mycenaeans | RB13 | 2 × 10⁻⁵ | rejected | 351,550 | SP 35.9 (2.3), CR2 41.2 (9.7), MY 22.9 (11.4) | |
Thessaloniki | Slavs + CR2 + Mycenaeans | RB13 | 1 × 10⁻⁵ | rejected | 351,550 | SP 43.7 (2.0), CR2 23.2 (8.8), MY 33.2 (10.2) | |
Athens | Slavs + CR3 + Mycenaeans | RB13 | 3 × 10⁻⁷ | rejected | 378,396 | SP 35.3 (2.0), CR3 44.3 (11.5), MY 20.5 (12.7) | |
Thessaloniki | Slavs + CR3 + Mycenaeans | RB13 | 3 × 10⁻⁶ | rejected | 378,396 | SP 43.9 (1.7), CR3 26.7 (8.1), MY 29.4 (9.0) | |
Athens | Slavs + CR4 + Mycenaeans | RB13 | 5 × 10⁻¹⁰ | rejected | 373,179 | SP 33.4 (1.8), CR4 27.7 (6.7), MY 39.0 (7.8) | |
Thessaloniki | Slavs + CR4 + Mycenaeans | RB13 | 7 × 10⁻⁷ | rejected | 373,179 | SP 43.2 (1.8), CR4 18.9 (5.3), MY 37.9 (6.3) | |
Athens | Slavs + CQ1 | R10 | 0.002 | rejected | 401,070 | SP 32.0 (2.8), CQ1 68.0 (2.8) | |
Thessaloniki | Slavs + CQ1 | R10 | 5 × 10⁻⁶ | rejected | 401,070 | SP 40.2 (2.7), CQ1 59.8 (2.7) | |
Athens | Slavs + CQ2 | R10 | 0.033 | rejected | 402,776 | SP 32.0 (3.0), CQ2 68.0 (3.0) | |
Thessaloniki | Slavs + CQ2 | R10 | 0.013 | rejected | 402,776 | SP 41.8 (2.9), CQ2 58.2 (2.9) | |
Athens | Slavs + CQ3 | R10 | 0.199 | fit | 407,319 | SP 34.0 (2.6), CQ3 66.0 (2.6) | |
Thessaloniki | Slavs + CQ3 | R10 | 0.068 | fit | 407,319 | SP 41.3 (2.6), CQ3 58.7 (2.6) | |
Athens | Slavs + CQ4 | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,998 | SP 332.9 (43.4), CQ4 −232.9 (43.4) | |
Thessaloniki | Slavs + CQ4 | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,998 | SP 281.3 (33.1), CQ4 −181.3 (33.1) | |
Athens | Slavs + CR1 | R10 | 0.069 | fit | 363,488 | SP 40.7 (2.3), CR1 59.3 (2.3) | |
Thessaloniki | Slavs + CR1 | R10 | 0.034 | rejected | 363,488 | SP 48.9 (2.2), CR1 51.1 (2.2) | |
Athens | Slavs + CR2 | R10 | 3 × 10⁻⁴ | rejected | 380,371 | SP 37.8 (2.7), CR2 62.2 (2.7) | |
Thessaloniki | Slavs + CR2 | R10 | 3 × 10⁻⁶ | rejected | 380,371 | SP 45.9 (2.5), CR2 54.1 (2.5) | |
Athens | Slavs + CR3 | R10 | 0.003 | rejected | 405,549 | SP 32.8 (2.7), CR3 67.2 (2.7) | |
Thessaloniki | Slavs + CR3 | R10 | 1 × 10⁻⁵ | rejected | 405,549 | SP 41.3 (2.7), CR3 58.7 (2.7) | |
Athens | Slavs + CR4 | R10 | 4 × 10⁻⁷ | rejected | 400,500 | SP 38.5 (2.4), CR4 61.5 (2.4) | |
Thessaloniki | Slavs + CR4 | R10 | 1 × 10⁻⁹ | rejected | 400,500 | SP 46.8 (2.4), CR4 53.2 (2.4) |
Members of the pairs in Table 14. allSnps false.
V6: raw counts. With the raw counts of Turkey_N (68) and Crete (34) instead of 43 and 5, Slavs + Kaman + Mycenaeans is rejected for Athens (p = 0.037; Mycenaeans 17.5 %, SE 7.2) and fits for Thessaloniki (p = 0.155; Table 7). V6 therefore does not hold. The grid of block 1 was not repeated on raw counts.
Verdict by the pre-declared rule: the Kaman three-source fits are INVALID. Two conditions of the rule are met: V1 is fragile for both targets and V6 does not hold (Athens). One such failure is decisive in the rule.
What this means. The fits Slavs + Old Hittite (Kaman) + Mycenaeans are not used as evidence for any statement about Greeks (the Mycenaean question, the claim that Old Hittite fits where Alalakh and Arslantepe do not). What it does not mean. It does not show the model to be false, and it does not show that Anatolian or Mycenaean ancestry is absent. The checks do not all point one way: V3 (Mycenaean source required), V4 (a second Kaman-site group fits) and V5 (no control pair passes) speak for the fit; V1 and V6 against it. The failures are narrow (p = 0.029 to 0.040), so the borderline calls are not stable under a small change of the data. Table 12 gives the results of the checks.
Table 12 | Results of the checks V1–V6
Check | Question | Athens (n=17) | Thessaloniki (n=14) |
|---|---|---|---|
V1 | remove one of the two persons | fragile: without MA2203 passes (p 0.348), without MA2200 rejected (p 0.040) | fragile: p 0.314 and 0.029 |
V2 | drop one right or change the panel (14 variants) | mixed: not rejected in 11 of 14 (rejected: Ukraine_EIA_Scythian dropped, R10, R14) | stable: not rejected in 12 of 14 (rejected: R10, R14) |
V3 | is the Mycenaean source needed? | required: Slavs + Kaman without MY rejected in 12 of 14 variants, drop-source p 0.018 | required: rejected in 14 of 14, drop-source p 0.006 |
V4 | does another, independent Kaman-site group also fit? | corroborated: Iron Age Kaman (n=2) fits, p 0.173, but Mycenaeans 50.2 % (not 21 %) | corroborated: p 0.146, Mycenaeans 46.9 % |
V5 | do generic pairs of persons also pass? | discriminating: 0 of 8 control pairs pass | discriminating: 0 of 8 |
V6 | raw instead of reduced counts of two right populations | does not hold: p 0.037 | holds: p 0.155 |
K8 | (descriptive) pairs passing without Mycenaeans on R10 | 2 of 8 | 1 of 8 |
Which statements stand. Table 13 sets the statements that circulate against the evidence of this study. A statement is called supported only for the sources that were tested.
Table 13 | Statements and evidence
Statement | Evidence in this study | Status |
|---|---|---|
A Slavic-type source is needed in every model | 0 of 34 models without it fit; all strict fits contain it | supported (tested sources) |
An Anatolian-type source is needed in every model | Slavs + Mycenaeans rejected on all four cells, p at most 3 × 10⁻¹¹ | supported (tested sources) |
A Mycenaean source is needed in every model | 10 strict two-source fits without it (small groups; six turn infeasible with Mycenaeans). Needed only in the Kaman model, which is INVALID | not supported; absence of Aegean ancestry not shown |
The Aegean share is well over 35 % | Passing models with a Kaman-site source: 5.3 to 50.2 % (12 models); Old Hittite fits 21.5 and 22.8 % (invalid); feasible three-source models 0.7 to 65 % | not identified |
Old Hittite fits where Alalakh and Arslantepe do not | Alalakh and Arslantepe are rejected (p ≤ 4 × 10⁻⁵); the Old Hittite fit is INVALID | not used as evidence |
Roman-era Anatolians + Slavs + Mycenaeans fit | 0 of 18 cells per target (9 Roman-era groups, 2 panels); the six cells with p ≥ 0.05 are infeasible | not supported |
The idea that the Kaman fits rest on one source. The objection that the fits come from one source of two persons is supported in the sense of fragility. The fit loses its better-covered person (V1), depends on two right counts (V6) and on the panel (V2). It is not supported in the sense that any two persons would fit (V5), and a second Kaman-site group can also produce a fit with another composition (V4). The Aegean share is not identified by these fits.
A qpAdm fit is the absence of rejection with the power of the data, not evidence of descent. The Slavic weight is the weight of a proxy; its specificity against other northern sources was not shown.
The posted models are replicated with proxies; tool, SNP set, individuals and right populations differ. M1–M4 use allSnps true, which distorts the models.
Several sources are small: Kaman (n=2), Aydın (n=1, 82,389 SNPs), Roman-era QW2 (n=2). Single persons and pairs have wide standard errors and pass more easily.
191 plus 116 single models and 4 sensitivity analyses (about 310 analyses) were run without a correction for multiplicity. The decisive p-values lie close to 0.05 (0.029, 0.034, 0.037, 0.040, 0.047).
The control pairs come from pools rejected as whole groups, not from groups known to be unrelated to Greeks.
The grid was not repeated on raw counts; another Aegean source in place of the Mycenaeans was not tried; why the pair passes more easily than its member and which right population causes the panel dependence were not tested.
Any further control after this block is a new, post hoc block and would need its own pre-declaration.
In these data, models need a Slavic-type and an Anatolian-type source to come near a fit; none of the tested models without either fits.
A Mycenaean source is not needed in every model: 10 two-source models without it are strict fits, all with small groups. This does not show that Aegean ancestry is absent.
Of the posted models, only the Old Hittite (Kaman) model fits (with other weights); the Roman-era, Alalakh and Arslantepe models are rejected.
The Kaman three-source fits are INVALID by the pre-declared rule (fragile without one person; not holding on raw counts) and are not used as evidence about Greeks.
The Mycenaean share is not identified: it depends on the source and on the panel.
Data and reproducibility. All models are stored as Nexju analyses; the appendices list every model with its p-value, weights and a link to its public result page. The pre-declared design and the results are in the Nexju project “nexju” of the author.
Haak, W. et al. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature 522, 207–211 (2015).
Maier, R. et al. On the limits of fitting complex models of population history to f-statistics. eLife 12, e85492 (2023).
Mallick, S. et al. The Allen Ancient DNA Resource (AADR): a curated compendium of ancient human genomes. Scientific Data 11, 182 (2024).
Table 14 | Groups used in the models
Code | Study name (n) | Nexju label / members |
|---|---|---|
SP | Poland Early Medieval Slavs | Poland_EarlyMedieval_Slav |
CZ | Czechia Early Medieval Early Slavs (n=3) | Czechia_EarlyMedieval_EarlySlav |
MY | Mycenaeans (n=7) | Mycenaean_HQ7 |
KAM2 | Old Hittite, Kaman (n=2) | Turkey_OldHittitePeriod |
K00 | Kaman person MA2200 (n=1) | Turkey_OldHittite_MA2200 |
K03 | Kaman person MA2203 (n=1) | Turkey_OldHittite_MA2203 |
AC3 | Kaman Assyrian Colony (n=3) | Turkey_Kaman_AssyrianColony3 |
KIA2 | Kaman Iron Age (n=2) | Turkey_Kaman_IronAge2 |
KAL7 | Kaman all periods (n=7) | Turkey_Kaman_Site7 |
OYL6 | Oylum Höyük MBA (n=6) | Turkey_Oylum_MBA6 |
HAR3 | Harmanoğlu–Göndürle EBA (n=3) | Turkey_Harmanoeren_EBA3 |
Q10 | Turkey Roman-era (n=10) | Turkey_Roman_Q10 |
QI5 | Turkey Roman-era QI5 (n=5) | Turkey_Roman_QI5 |
QL5 | Turkey Roman-era QL5 (n=5) | Turkey_Roman_QL5 |
QP3 | Turkey Roman-era QP3 (n=3) | Turkey_Roman_QP3 |
QB3 | Turkey Roman-era QB3 (n=3) | Turkey_Roman_QB3 |
QW2 | Turkey Roman-era QW2 (n=2) | Turkey_Roman_QW2 |
QS2 | Turkey Roman-era QS2 (n=2) | Turkey_Roman_QS2 |
QPW5 | Turkey Roman-era QPW5 (n=5) | Turkey_Roman_QPW5 |
O4 | Turkey Imperial Roman (n=4) | Turkey_ImperialRoman |
BE | Byzantine West Anatolia early (n=21) | Byz_West_Early |
BL | Byzantine West Anatolia late (n=8) | Byz_West_Late |
OTT5 | Turkey Ottoman (n=5) | Turkey_OttomanPeriod |
ALA5 | Alalakh MLBA (n=5) | Turkey_MLBA_Alalakh5 |
ARS16 | Arslantepe LC–EBA (n=16) | Turkey_LCEBA_Arslantepe16 |
WAB2 | Yassitepe BA (n=2) | Turkey_Aegean_YassitepeBA2 |
AYD1 | Aydın Bronze Age (n=1) | Turkey_Aegean_Aydin_BA |
LAZ4 | Lazarides LBA (n=4) | Greece_Lazarides_LBA |
LAZ1 | Lazarides EBA (n=1) | Greece_Lazarides_EBA |
LEB | Lebanon MBA (n=6) | Lebanon_MBA |
CQ1 | control pair (n=2) | NEV020, I4530 |
CQ2 | control pair (n=2) | BOG024, BOG020 |
CQ3 | control pair (n=2) | I8372, I20001 |
CQ4 | control pair (n=2) | I4529, I4531 |
CR1 | control pair (n=2) | ART014, ART005 |
CR2 | control pair (n=2) | ART019, ART026 |
CR3 | control pair (n=2) | ART032, ART010 |
CR4 | control pair (n=2) | ART027, ART022 |
Class: KA = single Kaman person, KZ = raw counts, KS = other Kaman-site and Anatolian sources, KN = control pairs; 3 = with Mycenaeans, 2 = without. Target A Athens (n=17), T Thessaloniki (n=14). Weights in percent (SE). All allSnps false. The link opens the public result page of the model. Codes as in Table 14.
Table 15 | The Kaman block, models 1–58 of 116
Cl. | T | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
KA3 | A | RB13 | 0.040 | rejected | 359,242 | SP 32.1 (2.2), K00 47.7 (7.6), MY 20.2 (8.1) | |
KA3 | T | RB13 | 0.029 | rejected | 359,242 | SP 42.2 (2.0), K00 35.9 (7.6), MY 21.9 (8.0) | |
KA3 | A | RB13 | 0.348 | fit | 114,164 | SP 30.2 (2.9), K03 46.8 (8.3), MY 23.0 (8.6) | |
KA3 | T | RB13 | 0.314 | fit | 114,164 | SP 39.7 (2.7), K03 34.3 (8.5), MY 26.0 (8.7) | |
KA2 | A | RB13 | 0.014 | rejected | 362,904 | SP 33.7 (2.6), K00 66.3 (2.6) | |
KA2 | T | RB13 | 0.007 | rejected | 362,904 | SP 43.9 (2.5), K00 56.1 (2.5) | |
KA2 | A | RB13 | 0.128 | fit | 115,179 | SP 32.1 (3.8), K03 67.9 (3.8) | |
KA2 | T | RB13 | 0.062 | fit | 115,179 | SP 41.6 (3.6), K03 58.4 (3.6) | |
KZ3 | A | RB13 | 0.037 | rejected | 371,818 | SP 32.7 (2.0), KAM2 49.9 (6.8), MY 17.5 (7.2) | |
KZ3 | T | RB13 | 0.155 | fit | 371,818 | SP 42.5 (1.8), KAM2 34.3 (6.1), MY 23.2 (6.3) | |
KZ2 | A | RB13 | 0.013 | rejected | 375,649 | SP 34.0 (2.2), KAM2 66.0 (2.2) | |
KZ2 | T | RB13 | 0.010 | rejected | 375,649 | SP 44.0 (2.2), KAM2 56.0 (2.2) | |
KS3 | A | R10 | 0.122 | fit | 318,702 | SP 37.2 (2.6), AC3 57.5 (6.8), MY 5.3 (7.5) | |
KS3 | A | RB13 | 0.014 | rejected | 303,961 | SP 32.7 (1.8), AC3 42.4 (5.8), MY 24.9 (6.5) | |
KS3 | T | R10 | 0.065 | fit | 318,702 | SP 45.0 (2.4), AC3 43.8 (6.0), MY 11.2 (6.5) | |
KS3 | T | RB13 | 0.037 | rejected | 303,961 | SP 43.2 (1.8), AC3 32.7 (5.5), MY 24.0 (6.1) | |
KS3 | A | R10 | 3 × 10⁻⁷ | rejected (infeasible) | 418,542 | SP 40.2 (3.8), HAR3 94.4 (21.2), MY −34.6 (23.2) | |
KS3 | A | RB13 | 4 × 10⁻⁹ | rejected (infeasible) | 395,104 | SP 61.0 (25.0), HAR3 238.9 (201.7), MY −199.9 (225.8) | |
KS3 | T | R10 | 7 × 10⁻⁹ | rejected (infeasible) | 418,542 | SP 46.2 (3.4), HAR3 65.9 (26.3), MY −12.1 (28.3) | |
KS3 | T | RB13 | 2 × 10⁻⁹ | rejected (infeasible) | 395,104 | SP −0.7 (259.4), HAR3 −359.3 (2172.3), MY 460.0 (2431.3) | |
KS3 | A | R10 | 0.064 | fit | 416,864 | SP 29.7 (2.0), KAL7 53.0 (4.9), MY 17.2 (4.7) | |
KS3 | A | RB13 | 0.052 | fit | 396,446 | SP 28.6 (1.4), KAL7 41.7 (4.4), MY 29.7 (4.4) | |
KS3 | T | R10 | 0.006 | rejected | 416,864 | SP 39.3 (2.0), KAL7 39.6 (4.8), MY 21.1 (4.6) | |
KS3 | T | RB13 | 0.024 | rejected | 396,446 | SP 40.0 (1.5), KAL7 31.4 (4.4), MY 28.6 (4.6) | |
KS3 | A | R10 | 0.033 | rejected | 203,370 | SP 12.2 (4.1), KIA2 38.7 (5.0), MY 49.1 (3.4) | |
KS3 | A | RB13 | 0.173 | fit | 194,713 | SP 17.2 (2.8), KIA2 32.6 (5.1), MY 50.2 (3.6) | |
KS3 | T | R10 | 0.084 | fit | 203,370 | SP 27.8 (3.9), KIA2 28.4 (4.8), MY 43.8 (3.0) | |
KS3 | T | RB13 | 0.146 | fit | 194,713 | SP 33.2 (2.8), KIA2 19.9 (4.7), MY 46.9 (3.1) | |
KS3 | A | R10 | 1 × 10⁻⁸ | rejected | 419,646 | SP 37.1 (1.8), OYL6 39.8 (4.3), MY 23.1 (4.9) | |
KS3 | A | RB13 | 1 × 10⁻⁵ | rejected | 398,364 | SP 34.9 (1.5), OYL6 32.6 (4.1), MY 32.5 (5.0) | |
KS3 | T | R10 | 1 × 10⁻⁶ | rejected | 419,646 | SP 44.3 (1.8), OYL6 26.9 (3.6), MY 28.8 (4.1) | |
KS3 | T | RB13 | 8 × 10⁻⁵ | rejected | 398,364 | SP 44.2 (1.6), OYL6 22.6 (3.8), MY 33.2 (4.6) | |
KS2 | A | R10 | 0.156 | fit | 320,567 | SP 37.9 (2.5), AC3 62.1 (2.5) | |
KS2 | A | RB13 | 6 × 10⁻⁴ | rejected | 307,060 | SP 35.9 (2.0), AC3 64.1 (2.0) | |
KS2 | T | R10 | 0.035 | rejected | 320,567 | SP 46.4 (2.5), AC3 53.6 (2.5) | |
KS2 | T | RB13 | 0.001 | rejected | 307,060 | SP 46.5 (1.9), AC3 53.5 (1.9) | |
KS2 | A | R10 | 9 × 10⁻⁹ | rejected | 420,614 | SP 36.7 (2.3), HAR3 63.3 (2.3) | |
KS2 | A | RB13 | 7 × 10⁻¹⁴ | rejected | 398,322 | SP 38.4 (1.6), HAR3 61.6 (1.6) | |
KS2 | T | R10 | 7 × 10⁻⁹ | rejected | 420,614 | SP 45.2 (2.4), HAR3 54.8 (2.4) | |
KS2 | T | RB13 | 5 × 10⁻¹¹ | rejected | 398,322 | SP 48.7 (1.6), HAR3 51.3 (1.6) | |
KS2 | A | R10 | 0.002 | rejected | 419,308 | SP 30.5 (2.4), KAL7 69.5 (2.4) | |
KS2 | A | RB13 | 2 × 10⁻⁷ | rejected | 400,484 | SP 30.1 (1.8), KAL7 69.9 (1.8) | |
KS2 | T | R10 | 2 × 10⁻⁶ | rejected | 419,308 | SP 40.1 (2.4), KAL7 59.9 (2.4) | |
KS2 | T | RB13 | 5 × 10⁻⁸ | rejected | 400,484 | SP 42.0 (1.8), KAL7 58.0 (1.8) | |
KS2 | A | R10 | 1 × 10⁻¹³ | rejected (infeasible) | 204,564 | SP −16.8 (12.4), KIA2 116.8 (12.4) | |
KS2 | A | RB13 | 8 × 10⁻¹⁰ | rejected (infeasible) | 196,653 | SP −10.5 (8.2), KIA2 110.5 (8.2) | |
KS2 | T | R10 | 8 × 10⁻¹⁵ | rejected (infeasible) | 204,564 | SP −0.7 (11.4), KIA2 100.7 (11.4) | |
KS2 | T | RB13 | 5 × 10⁻¹² | rejected | 196,653 | SP 4.5 (7.9), KIA2 95.5 (7.9) | |
KS2 | A | R10 | 1 × 10⁻¹³ | rejected | 422,002 | SP 41.5 (1.8), OYL6 58.5 (1.8) | |
KS2 | A | RB13 | 6 × 10⁻¹⁴ | rejected | 402,224 | SP 41.4 (1.2), OYL6 58.6 (1.2) | |
KS2 | T | R10 | < 1 × 10⁻¹⁶ | rejected | 422,002 | SP 50.6 (1.9), OYL6 49.4 (1.9) | |
KS2 | T | RB13 | 3 × 10⁻¹⁴ | rejected | 402,224 | SP 50.8 (1.4), OYL6 49.2 (1.4) | |
KN3 | A | R10 | 0.002 | rejected | 399,652 | SP 31.8 (2.6), CQ1 60.8 (9.5), MY 7.4 (9.2) | |
KN3 | A | RB13 | 8 × 10⁻⁵ | rejected | 373,765 | SP 31.5 (1.7), CQ1 45.5 (8.8), MY 23.1 (9.1) | |
KN3 | T | R10 | 1 × 10⁻⁵ | rejected | 399,652 | SP 40.1 (2.2), CQ1 43.3 (11.2), MY 16.6 (10.6) | |
KN3 | T | RB13 | 8 × 10⁻⁵ | rejected | 373,765 | SP 41.4 (1.5), CQ1 29.9 (8.1), MY 28.7 (8.5) | |
KN3 | A | R10 | 0.032 | rejected (infeasible) | 400,466 | SP 32.3 (3.3), CQ2 72.6 (12.4), MY −4.9 (12.6) | |
KN3 | A | RB13 | 0.020 | rejected | 380,744 | SP 32.8 (2.3), CQ2 62.8 (12.0), MY 4.4 (12.7) |
A Athens, T Thessaloniki. Classes as in the note above.
Table 16 | The Kaman block, models 59–116 of 116
Cl. | T | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
KN3 | T | R10 | 0.017 | rejected | 400,466 | SP 41.6 (2.8), CQ2 54.5 (11.4), MY 3.9 (11.4) | |
KN3 | T | RB13 | 0.008 | rejected | 380,744 | SP 43.0 (2.1), CQ2 45.1 (11.0), MY 11.9 (11.6) | |
KN3 | A | R10 | 0.265 | infeasible | 405,779 | SP 34.7 (3.1), CQ3 79.8 (12.0), MY −14.6 (12.2) | |
KN3 | A | RB13 | 9 × 10⁻⁶ | rejected | 379,530 | SP 34.9 (1.7), CQ3 34.7 (6.6), MY 30.4 (7.5) | |
KN3 | T | R10 | 0.059 | infeasible | 405,779 | SP 41.4 (2.7), CQ3 62.3 (11.6), MY −3.7 (11.5) | |
KN3 | T | RB13 | 9 × 10⁻⁵ | rejected | 379,530 | SP 43.3 (1.6), CQ3 23.4 (5.6), MY 33.3 (6.3) | |
KN3 | A | R10 | 7 × 10⁻⁶ | rejected (infeasible) | 422,515 | SP −25.2 (9.3), CQ4 47.9 (6.6), MY 77.3 (3.6) | |
KN3 | A | RB13 | 1 × 10⁻¹⁵ | rejected | 401,630 | SP 21.7 (3.5), CQ4 9.8 (3.2), MY 68.5 (1.7) | |
KN3 | T | R10 | 4 × 10⁻⁵ | rejected | 422,515 | SP 0.9 (8.4), CQ4 33.7 (5.9), MY 65.4 (3.2) | |
KN3 | T | RB13 | 6 × 10⁻¹¹ | rejected | 401,630 | SP 37.1 (3.4), CQ4 4.7 (3.1), MY 58.2 (1.7) | |
KN3 | A | R10 | 0.170 | infeasible | 362,635 | SP 44.5 (3.7), CR1 75.7 (12.4), MY −20.2 (14.4) | |
KN3 | A | RB13 | 8 × 10⁻⁵ | rejected (infeasible) | 335,134 | SP 42.7 (3.6), CR1 65.8 (16.7), MY −8.5 (19.7) | |
KN3 | T | R10 | 0.042 | rejected (infeasible) | 362,635 | SP 50.9 (3.2), CR1 59.5 (11.8), MY −10.3 (13.6) | |
KN3 | T | RB13 | 2 × 10⁻⁴ | rejected | 335,134 | SP 49.7 (2.9), CR1 44.7 (12.4), MY 5.6 (14.7) | |
KN3 | A | R10 | 9 × 10⁻⁴ | rejected | 379,319 | SP 36.1 (2.4), CR2 49.2 (6.8), MY 14.7 (7.4) | |
KN3 | A | RB13 | 2 × 10⁻⁵ | rejected | 351,550 | SP 35.9 (2.3), CR2 41.2 (9.7), MY 22.9 (11.4) | |
KN3 | T | R10 | 8 × 10⁻⁵ | rejected | 379,319 | SP 43.4 (2.2), CR2 35.9 (6.7), MY 20.6 (7.2) | |
KN3 | T | RB13 | 1 × 10⁻⁵ | rejected | 351,550 | SP 43.7 (2.0), CR2 23.2 (8.8), MY 33.2 (10.2) | |
KN3 | A | R10 | 0.007 | rejected | 404,033 | SP 32.0 (2.4), CR3 54.9 (7.5), MY 13.1 (7.5) | |
KN3 | A | RB13 | 3 × 10⁻⁷ | rejected | 378,396 | SP 35.3 (2.0), CR3 44.3 (11.5), MY 20.5 (12.7) | |
KN3 | T | R10 | 2 × 10⁻⁴ | rejected | 404,033 | SP 40.7 (2.2), CR3 41.5 (7.4), MY 17.8 (7.3) | |
KN3 | T | RB13 | 3 × 10⁻⁶ | rejected | 378,396 | SP 43.9 (1.7), CR3 26.7 (8.1), MY 29.4 (9.0) | |
KN3 | A | R10 | 2 × 10⁻⁷ | rejected | 399,125 | SP 37.2 (2.7), CR4 53.3 (8.9), MY 9.5 (10.1) | |
KN3 | A | RB13 | 5 × 10⁻¹⁰ | rejected | 373,179 | SP 33.4 (1.8), CR4 27.7 (6.7), MY 39.0 (7.8) | |
KN3 | T | R10 | 4 × 10⁻⁸ | rejected | 399,125 | SP 44.3 (2.1), CR4 31.8 (7.4), MY 23.9 (7.8) | |
KN3 | T | RB13 | 7 × 10⁻⁷ | rejected | 373,179 | SP 43.2 (1.8), CR4 18.9 (5.3), MY 37.9 (6.3) | |
KN2 | A | R10 | 0.002 | rejected | 401,070 | SP 32.0 (2.8), CQ1 68.0 (2.8) | |
KN2 | A | RB13 | 6 × 10⁻⁶ | rejected | 375,948 | SP 32.8 (2.1), CQ1 67.2 (2.1) | |
KN2 | T | R10 | 5 × 10⁻⁶ | rejected | 401,070 | SP 40.2 (2.7), CQ1 59.8 (2.7) | |
KN2 | T | RB13 | 6 × 10⁻⁷ | rejected | 375,948 | SP 43.0 (1.9), CQ1 57.0 (1.9) | |
KN2 | A | R10 | 0.033 | rejected | 402,776 | SP 32.0 (3.0), CQ2 68.0 (3.0) | |
KN2 | A | RB13 | 0.041 | rejected | 384,556 | SP 33.2 (2.2), CQ2 66.8 (2.2) | |
KN2 | T | R10 | 0.013 | rejected | 402,776 | SP 41.8 (2.9), CQ2 58.2 (2.9) | |
KN2 | T | RB13 | 0.013 | rejected | 384,556 | SP 44.0 (2.2), CQ2 56.0 (2.2) | |
KN2 | A | R10 | 0.199 | fit | 407,319 | SP 34.0 (2.6), CQ3 66.0 (2.6) | |
KN2 | A | RB13 | 5 × 10⁻⁹ | rejected | 381,874 | SP 39.0 (1.8), CQ3 61.0 (1.8) | |
KN2 | T | R10 | 0.068 | fit | 407,319 | SP 41.3 (2.6), CQ3 58.7 (2.6) | |
KN2 | T | RB13 | 5 × 10⁻¹⁰ | rejected | 381,874 | SP 47.7 (1.8), CQ3 52.3 (1.8) | |
KN2 | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,998 | SP 332.9 (43.4), CQ4 −232.9 (43.4) | |
KN2 | A | RB13 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 405,731 | SP 1739.0 (1644.3), CQ4 −1639.0 (1644.3) | |
KN2 | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,998 | SP 281.3 (33.1), CQ4 −181.3 (33.1) | |
KN2 | T | RB13 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 405,731 | SP 751.0 (325.9), CQ4 −651.0 (325.9) | |
KN2 | A | R10 | 0.069 | fit | 363,488 | SP 40.7 (2.3), CR1 59.3 (2.3) | |
KN2 | A | RB13 | 1 × 10⁻⁴ | rejected | 336,417 | SP 41.4 (1.7), CR1 58.6 (1.7) | |
KN2 | T | R10 | 0.034 | rejected | 363,488 | SP 48.9 (2.2), CR1 51.1 (2.2) | |
KN2 | T | RB13 | 3 × 10⁻⁴ | rejected | 336,417 | SP 50.7 (1.6), CR1 49.3 (1.6) | |
KN2 | A | R10 | 3 × 10⁻⁴ | rejected | 380,371 | SP 37.8 (2.7), CR2 62.2 (2.7) | |
KN2 | A | RB13 | 5 × 10⁻⁶ | rejected | 353,158 | SP 39.4 (1.8), CR2 60.6 (1.8) | |
KN2 | T | R10 | 3 × 10⁻⁶ | rejected | 380,371 | SP 45.9 (2.5), CR2 54.1 (2.5) | |
KN2 | T | RB13 | 9 × 10⁻⁸ | rejected | 353,158 | SP 48.7 (1.8), CR2 51.3 (1.8) | |
KN2 | A | R10 | 0.003 | rejected | 405,549 | SP 32.8 (2.7), CR3 67.2 (2.7) | |
KN2 | A | RB13 | 1 × 10⁻⁸ | rejected | 380,748 | SP 37.5 (1.8), CR3 62.5 (1.8) | |
KN2 | T | R10 | 1 × 10⁻⁵ | rejected | 405,549 | SP 41.3 (2.7), CR3 58.7 (2.7) | |
KN2 | T | RB13 | 4 × 10⁻⁹ | rejected | 380,748 | SP 47.4 (1.8), CR3 52.6 (1.8) | |
KN2 | A | R10 | 4 × 10⁻⁷ | rejected | 400,500 | SP 38.5 (2.4), CR4 61.5 (2.4) | |
KN2 | A | RB13 | 5 × 10⁻¹⁴ | rejected | 375,269 | SP 39.7 (1.7), CR4 60.3 (1.7) | |
KN2 | T | R10 | 1 × 10⁻⁹ | rejected | 400,500 | SP 46.8 (2.4), CR4 53.2 (2.4) | |
KN2 | T | RB13 | 1 × 10⁻¹² | rejected | 375,269 | SP 49.8 (1.7), CR4 50.2 (1.7) |
A Athens, T Thessaloniki. Classes as in the note above.
The right-panel variants are the 12 drop-one-right variants and the panels R10 and R14. Mycenaean weight (MY) in percent where the model has Mycenaeans. The four analyses are not single models and have no public link; each is identified by its analysis id. All allSnps false.
Table 17 | Slavs + Kaman + Mycenaeans, Modern Greeks (Athens, n=17)
No. | Variant | p | Result | SNPs | MY % |
|---|---|---|---|---|---|
1 | base panel RB13 | 0.128 | not rejected | 369,507 | 21.5 |
2 | drop right: Russia_Samara_EBA_Yamnaya | 0.094 | not rejected | 369,507 | 21.0 |
3 | drop right: Serbia_IronGates_Mesolithic | 0.235 | not rejected | 369,507 | 21.1 |
4 | drop right: Russia_Karelia_Mesolithic | 0.116 | not rejected | 373,836 | 21.3 |
5 | drop right: Iran_GanjDareh_N | 0.137 | not rejected | 369,885 | 18.6 |
6 | drop right: Turkey_N | 0.129 | not rejected | 369,507 | 23.0 |
7 | drop right: Spain_IA_CeltIberian | 0.099 | not rejected | 369,508 | 21.6 |
8 | drop right: Greece_Crete_HgCharalambos_EMBA | 0.089 | not rejected | 371,821 | 21.2 |
9 | drop right: Croatia_MLBA | 0.175 | not rejected | 371,572 | 15.9 |
10 | drop right: Netherlands_MBA | 0.140 | not rejected | 371,887 | 23.1 |
11 | drop right: Ukraine_EIA_Scythian | 0.034 | rejected | 401,493 | 18.7 |
12 | drop right: Byz_SE_Early | 0.090 | not rejected | 377,661 | 21.2 |
13 | drop right: Estonia_LBA_StoneCistGraves | 0.226 | not rejected | 385,574 | 16.8 |
14 | panel R10 | 0.047 | rejected | 387,656 | 12.0 |
15 | panel R14 | 0.001 | rejected | 386,921 | 13.3 |
16 | drop source: Poland_EarlyMedieval_Slav | 1 × 10⁻¹⁵ | rejected (infeasible) | 417,424 | −241.8 |
17 | drop source: Turkey_OldHittitePeriod | < 1 × 10⁻¹⁶ | rejected | 401,633 | 69.6 |
18 | drop source: Mycenaean_HQ7 | 0.018 | rejected | 373,260 | – |
Analysis id 4582d8e2-b327-423f-a0cf-002e254e997f.
Table 18 | Slavs + Kaman + Mycenaeans, Modern Greeks (Thessaloniki, n=14)
No. | Variant | p | Result | SNPs | MY % |
|---|---|---|---|---|---|
1 | base panel RB13 | 0.083 | not rejected | 369,507 | 22.8 |
2 | drop right: Russia_Samara_EBA_Yamnaya | 0.056 | not rejected | 369,507 | 22.8 |
3 | drop right: Serbia_IronGates_Mesolithic | 0.232 | not rejected | 369,507 | 22.1 |
4 | drop right: Russia_Karelia_Mesolithic | 0.100 | not rejected | 373,836 | 23.3 |
5 | drop right: Iran_GanjDareh_N | 0.074 | not rejected | 369,885 | 20.2 |
6 | drop right: Turkey_N | 0.079 | not rejected | 369,507 | 24.6 |
7 | drop right: Spain_IA_CeltIberian | 0.070 | not rejected | 369,508 | 23.2 |
8 | drop right: Greece_Crete_HgCharalambos_EMBA | 0.153 | not rejected | 371,821 | 23.8 |
9 | drop right: Croatia_MLBA | 0.097 | not rejected | 371,572 | 18.6 |
10 | drop right: Netherlands_MBA | 0.236 | not rejected | 371,887 | 25.0 |
11 | drop right: Ukraine_EIA_Scythian | 0.066 | not rejected | 401,493 | 22.1 |
12 | drop right: Byz_SE_Early | 0.100 | not rejected | 377,661 | 21.6 |
13 | drop right: Estonia_LBA_StoneCistGraves | 0.096 | not rejected | 385,574 | 19.6 |
14 | panel R10 | 0.003 | rejected | 387,656 | 14.2 |
15 | panel R14 | 2 × 10⁻⁴ | rejected | 386,921 | 16.1 |
16 | drop source: Poland_EarlyMedieval_Slav | 1 × 10⁻¹⁶ | rejected (infeasible) | 417,424 | −1,219.1 |
17 | drop source: Turkey_OldHittitePeriod | 3 × 10⁻¹¹ | rejected | 401,633 | 58.6 |
18 | drop source: Mycenaean_HQ7 | 0.006 | rejected | 373,260 | – |
Analysis id 3da80129-486e-49f6-89aa-892930c2ab05.
Table 19 | Slavs + Kaman (no Mycenaean source), Modern Greeks (Athens, n=17)
No. | Variant | p | Result | SNPs |
|---|---|---|---|---|
1 | base panel RB13 | 0.018 | rejected | 373,260 |
2 | drop right: Russia_Samara_EBA_Yamnaya | 0.015 | rejected | 373,260 |
3 | drop right: Serbia_IronGates_Mesolithic | 0.039 | rejected | 373,260 |
4 | drop right: Russia_Karelia_Mesolithic | 0.017 | rejected | 378,399 |
5 | drop right: Iran_GanjDareh_N | 0.049 | rejected | 373,686 |
6 | drop right: Turkey_N | 0.012 | rejected | 373,260 |
7 | drop right: Spain_IA_CeltIberian | 0.012 | rejected | 373,261 |
8 | drop right: Greece_Crete_HgCharalambos_EMBA | 0.012 | rejected | 375,654 |
9 | drop right: Croatia_MLBA | 0.097 | not rejected | 375,844 |
10 | drop right: Netherlands_MBA | 0.011 | rejected | 376,176 |
11 | drop right: Ukraine_EIA_Scythian | 0.013 | rejected | 405,618 |
12 | drop right: Byz_SE_Early | 0.010 | rejected | 382,698 |
13 | drop right: Estonia_LBA_StoneCistGraves | 0.125 | not rejected | 389,532 |
14 | panel R10 | 0.045 | rejected | 389,918 |
15 | panel R14 | 0.001 | rejected | 389,046 |
16 | drop source: Poland_EarlyMedieval_Slav | < 1 × 10⁻¹⁶ | rejected | 421,717 |
17 | drop source: Turkey_OldHittitePeriod | < 1 × 10⁻¹⁶ | rejected | 405,734 |
Analysis id b2bf9833-ace4-4ae5-91f6-d6cc85c8893e.
Table 20 | Slavs + Kaman (no Mycenaean source), Modern Greeks (Thessaloniki, n=14)
No. | Variant | p | Result | SNPs |
|---|---|---|---|---|
1 | base panel RB13 | 0.006 | rejected | 373,260 |
2 | drop right: Russia_Samara_EBA_Yamnaya | 0.004 | rejected | 373,260 |
3 | drop right: Serbia_IronGates_Mesolithic | 0.026 | rejected | 373,260 |
4 | drop right: Russia_Karelia_Mesolithic | 0.007 | rejected | 378,399 |
5 | drop right: Iran_GanjDareh_N | 0.019 | rejected | 373,686 |
6 | drop right: Turkey_N | 0.003 | rejected | 373,260 |
7 | drop right: Spain_IA_CeltIberian | 0.004 | rejected | 373,261 |
8 | drop right: Greece_Crete_HgCharalambos_EMBA | 0.008 | rejected | 375,654 |
9 | drop right: Croatia_MLBA | 0.031 | rejected | 375,844 |
10 | drop right: Netherlands_MBA | 0.007 | rejected | 376,176 |
11 | drop right: Ukraine_EIA_Scythian | 0.009 | rejected | 405,618 |
12 | drop right: Byz_SE_Early | 0.007 | rejected | 382,698 |
13 | drop right: Estonia_LBA_StoneCistGraves | 0.031 | rejected | 389,532 |
14 | panel R10 | 0.002 | rejected | 389,918 |
15 | panel R14 | 6 × 10⁻⁵ | rejected | 389,046 |
16 | drop source: Poland_EarlyMedieval_Slav | < 1 × 10⁻¹⁶ | rejected | 421,717 |
17 | drop source: Turkey_OldHittitePeriod | < 1 × 10⁻¹⁶ | rejected | 405,734 |
Analysis id d3fcbc4d-d25b-4381-9610-5b4072ba4ffc.
Classes: TP1–TP4 posted models M1–M4 (allSnps true, marked †, Thessaloniki), CA1–CA3 Antoniou models, RZ replications on raw counts, G3 Slavs + X + Mycenaeans, G2 Slavs + X, G2b X + Mycenaeans, G2c Slavs + Mycenaeans. allSnps true (†) stretches the models and distorts p-values; all other models are allSnps false. Target A Athens (n=17), T Thessaloniki (n=14). Codes as in Table 14.
Table 21 | Block 1, models 1–48 of 191
Cl. | T | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
TP1 | T | RB13 | 4 × 10⁻⁵ † | rejected (infeasible) | 468,442 | MY 73.0 (20.5), Q10 123.4 (58.3), SP 21.2 (9.8), OTT5 −117.6 (67.1) | |
TP2 | T | RB13 | 2 × 10⁻⁵ † | rejected | 468,442 | MY 26.5 (5.0), ALA5 28.4 (4.3), SP 45.0 (1.5) | |
TP3 | T | RB13 | 2 × 10⁻¹¹ † | rejected | 468,442 | MY 33.8 (6.3), ARS16 21.1 (5.1), SP 45.2 (1.8) | |
TP4 | T | RB13 | 0.071† | fit | 468,442 | MY 19.4 (6.0), KAM2 39.0 (5.8), SP 41.6 (1.8) | |
CA1 | T | RB13 | 0.653 | imprecise | 27,781 | LAZ4 11.4 (480.9), LEB 4.6 (68.2), CZ 34.7 (82.9), AYD1 49.3 (630.2) | |
CA1 | A | RB13 | 0.905 | imprecise | 27,781 | LAZ4 24.2 (19.1), LEB 16.1 (9.7), CZ 27.1 (5.5), AYD1 32.6 (23.4) | |
CA2 | T | RB13 | 0.959 | imprecise | 12,701 | LAZ1 13.8 (8.9), LEB 28.5 (16.6), CZ 43.0 (8.9), AYD1 14.7 (26.6) | |
CA2 | A | RB13 | 0.850 | imprecise | 12,701 | LAZ1 16.2 (16.3), LEB 44.6 (30.7), CZ 37.4 (17.8), AYD1 1.8 (60.1) | |
CA3 | T | RB13 | 0.838 | infeasible | 215,350 | LAZ4 884.2 (5213.5), LEB −102.9 (745.4), CZ −49.9 (553.7), YAS2 −631.4 (3920.3) | |
CA3 | A | RB13 | 0.845 | infeasible | 215,350 | LAZ4 −573.6 (3621.8), LEB 110.7 (514.0), CZ 96.2 (385.4), YAS2 466.7 (2726.7) | |
RZ | T | RB13 | 0.748 | imprecise | 27,903 | LAZ4 21.7 (33.3), LEB 9.2 (10.7), CZ 38.7 (6.5), AYD1 30.4 (38.7) | |
RZ | T | RB13 | 0.921 | imprecise | 12,736 | LAZ1 12.9 (10.4), LEB 34.2 (19.3), CZ 46.8 (9.9), AYD1 6.1 (32.8) | |
RZ | T | RB13 | 0.696 | infeasible | 216,059 | LAZ4 448.9 (1193.5), LEB −32.1 (144.7), CZ 0.0 (119.7), YAS2 −316.8 (935.7) | |
RZ | T | RB13 | 2 × 10⁻⁴ † | rejected (infeasible) | 468,442 | MY 69.9 (19.0), Q10 124.8 (55.6), SP 21.2 (9.5), OTT5 −115.9 (63.3) | |
RZ | T | RB13 | 4 × 10⁻⁵ † | rejected | 468,442 | MY 27.5 (4.6), ALA5 27.5 (4.0), SP 45.0 (1.5) | |
RZ | T | RB13 | 1 × 10⁻⁹ † | rejected | 468,442 | MY 30.2 (5.6), ARS16 23.7 (4.6), SP 46.1 (1.6) | |
RZ | T | RB13 | 0.056† | fit | 468,442 | MY 19.0 (6.1), KAM2 38.8 (5.9), SP 42.2 (1.8) | |
G3 | T | R10 | 5 × 10⁻⁶ | rejected | 419,215 | SP 43.9 (1.8), ALA5 31.0 (4.3), MY 25.1 (4.7) | |
G3 | T | R10 | 4 × 10⁻⁹ | rejected | 422,450 | SP 45.1 (1.8), ARS16 33.1 (5.3), MY 21.8 (5.8) | |
G3 | T | R10 | 0.339 | imprecise | 31,573 | SP 38.3 (9.7), AYD1 54.4 (29.5), MY 7.2 (23.8) | |
G3 | T | R10 | 3 × 10⁻⁶ | rejected | 422,446 | SP 34.2 (2.1), BE 61.7 (8.8), MY 4.1 (7.6) | |
G3 | T | R10 | 2 × 10⁻⁷ | rejected | 419,083 | SP 36.5 (2.3), BL 55.0 (10.2), MY 8.5 (9.1) | |
G3 | T | R10 | 0.003 | rejected | 387,656 | SP 40.8 (2.6), KAM2 45.0 (8.7), MY 14.2 (8.5) | |
G3 | T | R10 | 0.497 | infeasible | 333,039 | SP 36.7 (3.3), O4 72.3 (15.4), MY −9.0 (14.3) | |
G3 | T | R10 | 8 × 10⁻⁶ | rejected | 418,180 | SP 38.1 (2.0), OTT5 39.5 (5.4), MY 22.3 (4.9) | |
G3 | T | R10 | 1 × 10⁻⁵ | rejected | 422,516 | SP 29.1 (2.6), Q10 41.0 (4.9), MY 29.9 (3.5) | |
G3 | T | R10 | 2 × 10⁻⁵ | rejected | 422,515 | SP 8.9 (6.3), QB3 38.1 (6.0), MY 52.9 (2.2) | |
G3 | T | R10 | 7 × 10⁻⁵ | rejected | 421,133 | SP 39.4 (2.0), QI5 41.1 (5.9), MY 19.5 (5.6) | |
G3 | T | R10 | 1 × 10⁻⁵ | rejected | 422,516 | SP 20.4 (4.0), QL5 41.6 (5.6), MY 38.0 (2.9) | |
G3 | T | R10 | 0.027 | rejected | 412,512 | SP 40.1 (2.5), QP3 54.7 (9.5), MY 5.1 (9.1) | |
G3 | T | R10 | 0.065 | infeasible | 421,738 | SP 40.6 (2.2), QPW5 60.4 (8.4), MY −1.0 (8.2) | |
G3 | T | R10 | 5 × 10⁻⁶ | rejected | 389,758 | SP 39.2 (2.2), QS2 29.6 (5.4), MY 31.2 (5.0) | |
G3 | T | R10 | 0.059 | infeasible | 405,779 | SP 41.4 (2.7), QW2 62.3 (11.6), MY −3.7 (11.5) | |
G3 | T | R10 | 0.001 | rejected (infeasible) | 299,546 | SP −299.8 (36779.1), WAB2 −2041.5 (220016.6), MY 2441.3 (256794.0) | |
G3 | T | RB13 | 4 × 10⁻⁵ | rejected | 397,369 | SP 44.0 (1.6), ALA5 25.2 (4.6), MY 30.7 (5.3) | |
G3 | T | RB13 | 3 × 10⁻⁹ | rejected | 401,507 | SP 44.4 (1.9), ARS16 19.3 (5.7), MY 36.3 (7.0) | |
G3 | T | RB13 | 0.859 | imprecise | 30,372 | SP 27.5 (5.3), AYD1 59.3 (18.1), MY 13.2 (15.9) | |
G3 | T | RB13 | 1 × 10⁻⁹ | rejected | 401,476 | SP 39.7 (1.4), BE 35.1 (15.1), MY 25.2 (14.8) | |
G3 | T | RB13 | 9 × 10⁻¹⁰ | rejected | 395,855 | SP 41.2 (1.5), BL 24.7 (11.0), MY 34.2 (11.3) | |
G3 | T | RB13 | 0.083 | fit | 369,507 | SP 41.8 (1.8), KAM2 35.3 (6.2), MY 22.8 (6.5) | |
G3 | T | RB13 | 0.004 | rejected | 306,512 | SP 40.5 (2.1), O4 55.3 (20.5), MY 4.1 (20.6) | |
G3 | T | RB13 | 7 × 10⁻⁹ | rejected | 394,423 | SP 40.5 (1.5), OTT5 19.5 (6.8), MY 40.0 (7.0) | |
G3 | T | RB13 | 4 × 10⁻⁸ | rejected | 401,633 | SP 37.9 (1.7), Q10 19.8 (4.8), MY 42.2 (4.4) | |
G3 | T | RB13 | 5 × 10⁻¹¹ | rejected | 401,631 | SP 37.5 (3.2), QB3 6.0 (4.1), MY 56.5 (2.2) | |
G3 | T | RB13 | 1 × 10⁻⁴ | rejected | 400,035 | SP 41.9 (1.5), QI5 29.8 (5.4), MY 28.3 (5.9) | |
G3 | T | RB13 | 2 × 10⁻⁹ | rejected | 401,633 | SP 36.6 (2.2), QL5 13.8 (4.4), MY 49.6 (3.4) | |
G3 | T | RB13 | 0.023 | rejected | 391,548 | SP 42.5 (1.9), QP3 48.5 (10.3), MY 9.0 (10.8) | |
G3 | T | RB13 | 0.005 | rejected | 400,574 | SP 43.5 (1.6), QPW5 38.0 (6.4), MY 18.5 (6.9) |
† allSnps true. It stretches the models and distorts p-values; these p-values are not comparable with allSnps false results.
Table 22 | Block 1, models 49–96 of 191
Cl. | T | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
G3 | T | RB13 | 8 × 10⁻⁶ | rejected | 365,500 | SP 41.5 (1.5), QS2 17.8 (4.3), MY 40.8 (4.8) | |
G3 | T | RB13 | 9 × 10⁻⁵ | rejected | 379,530 | SP 43.3 (1.6), QW2 23.4 (5.6), MY 33.3 (6.3) | |
G3 | T | RB13 | 0.010 | rejected (infeasible) | 269,625 | SP 20.6 (21.0), WAB2 −141.1 (140.1), MY 220.5 (160.5) | |
G3 | A | R10 | 8 × 10⁻⁷ | rejected | 419,215 | SP 36.8 (2.0), ALA5 46.0 (5.2), MY 17.2 (5.9) | |
G3 | A | R10 | 5 × 10⁻¹¹ | rejected | 422,450 | SP 38.7 (1.8), ARS16 50.5 (5.8), MY 10.8 (6.6) | |
G3 | A | R10 | 0.547 | imprecise | 31,573 | SP 29.1 (9.4), AYD1 65.6 (24.5), MY 5.3 (20.9) | |
G3 | A | R10 | 2 × 10⁻⁴ | rejected (infeasible) | 422,446 | SP 22.8 (1.9), BE 87.4 (7.8), MY −10.2 (7.0) | |
G3 | A | R10 | 9 × 10⁻⁶ | rejected (infeasible) | 419,083 | SP 25.1 (2.5), BL 82.5 (10.4), MY −7.6 (9.1) | |
G3 | A | R10 | 0.047 | rejected | 387,656 | SP 32.7 (2.7), KAM2 55.3 (8.1), MY 12.0 (8.1) | |
G3 | A | R10 | 0.396 | infeasible | 333,039 | SP 26.5 (4.1), O4 95.4 (19.0), MY −21.9 (17.5) | |
G3 | A | R10 | 8 × 10⁻⁷ | rejected | 418,180 | SP 28.4 (2.1), OTT5 56.1 (6.3), MY 15.5 (5.9) | |
G3 | A | R10 | 3 × 10⁻⁶ | rejected | 422,516 | SP 14.4 (2.7), Q10 59.8 (5.2), MY 25.8 (3.7) | |
G3 | A | R10 | 1 × 10⁻⁵ | rejected (infeasible) | 422,515 | SP −15.1 (6.7), QB3 55.5 (6.3), MY 59.7 (2.5) | |
G3 | A | R10 | 3 × 10⁻⁴ | rejected | 421,133 | SP 29.0 (2.2), QI5 58.6 (6.4), MY 12.3 (6.1) | |
G3 | A | R10 | 3 × 10⁻⁶ | rejected | 422,516 | SP 1.7 (4.3), QL5 61.6 (6.3), MY 36.7 (3.2) | |
G3 | A | R10 | 0.039 | rejected (infeasible) | 412,512 | SP 30.2 (2.8), QP3 73.6 (10.5), MY −3.8 (10.3) | |
G3 | A | R10 | 0.217 | infeasible | 421,738 | SP 32.0 (2.4), QPW5 79.7 (8.8), MY −11.7 (8.8) | |
G3 | A | R10 | 8 × 10⁻⁵ | rejected | 389,758 | SP 29.6 (2.4), QS2 44.0 (5.8), MY 26.4 (5.4) | |
G3 | A | R10 | 0.265 | infeasible | 405,779 | SP 34.7 (3.1), QW2 79.8 (12.0), MY −14.6 (12.2) | |
G3 | A | R10 | 0.003 | rejected (infeasible) | 299,546 | SP 79.4 (41.3), WAB2 275.7 (235.4), MY −255.1 (275.2) | |
G3 | A | RB13 | 9 × 10⁻⁶ | rejected | 397,369 | SP 34.9 (1.6), ALA5 36.3 (5.4), MY 28.8 (6.3) | |
G3 | A | RB13 | 2 × 10⁻¹³ | rejected | 401,507 | SP 35.4 (1.9), ARS16 28.3 (6.2), MY 36.2 (7.6) | |
G3 | A | RB13 | 0.714 | imprecise | 30,372 | SP 19.0 (6.4), AYD1 70.6 (21.8), MY 10.3 (18.6) | |
G3 | A | RB13 | 3 × 10⁻¹² | rejected (infeasible) | 401,476 | SP 27.7 (1.6), BE 82.6 (24.5), MY −10.3 (23.5) | |
G3 | A | RB13 | 3 × 10⁻¹⁴ | rejected | 395,855 | SP 30.1 (1.4), BL 46.6 (26.4), MY 23.4 (26.2) | |
G3 | A | RB13 | 0.128 | fit | 369,507 | SP 31.3 (1.9), KAM2 47.2 (6.3), MY 21.5 (6.7) | |
G3 | A | RB13 | 0.008 | rejected (infeasible) | 306,512 | SP 29.6 (2.9), O4 88.9 (25.1), MY −18.5 (25.4) | |
G3 | A | RB13 | 1 × 10⁻¹³ | rejected | 394,423 | SP 29.8 (1.4), OTT5 28.4 (9.0), MY 41.7 (9.2) | |
G3 | A | RB13 | 7 × 10⁻¹¹ | rejected | 401,633 | SP 26.0 (1.5), Q10 28.8 (4.8), MY 45.1 (4.4) | |
G3 | A | RB13 | 7 × 10⁻¹⁶ | rejected | 401,631 | SP 22.6 (3.3), QB3 12.4 (4.3), MY 65.0 (2.2) | |
G3 | A | RB13 | 3 × 10⁻⁵ | rejected | 400,035 | SP 31.8 (1.4), QI5 42.7 (6.1), MY 25.5 (6.4) | |
G3 | A | RB13 | 1 × 10⁻¹² | rejected | 401,633 | SP 23.4 (2.1), QL5 22.1 (4.6), MY 54.5 (3.5) | |
G3 | A | RB13 | 0.031 | rejected | 391,548 | SP 32.2 (2.0), QP3 67.1 (11.5), MY 0.7 (12.0) | |
G3 | A | RB13 | 0.024 | rejected | 400,574 | SP 34.3 (1.4), QPW5 53.6 (6.5), MY 12.1 (6.9) | |
G3 | A | RB13 | 4 × 10⁻⁶ | rejected | 365,500 | SP 31.4 (1.5), QS2 26.7 (4.2), MY 41.8 (4.7) | |
G3 | A | RB13 | 9 × 10⁻⁶ | rejected | 379,530 | SP 34.9 (1.7), QW2 34.7 (6.6), MY 30.4 (7.5) | |
G3 | A | RB13 | 0.004 | rejected (infeasible) | 269,625 | SP −94.5 (392.0), WAB2 −852.4 (2663.2), MY 1047.0 (3054.3) | |
G2 | T | R10 | 2 × 10⁻¹¹ | rejected | 421,441 | SP 48.4 (2.0), ALA5 51.6 (2.0) | |
G2 | T | R10 | 3 × 10⁻¹³ | rejected | 424,908 | SP 48.4 (1.7), ARS16 51.6 (1.7) | |
G2 | T | R10 | 0.448 | fit | 31,719 | SP 35.9 (10.0), AYD1 64.1 (10.0) | |
G2 | T | R10 | 4 × 10⁻⁶ | rejected | 424,898 | SP 33.9 (1.9), BE 66.1 (1.9) | |
G2 | T | R10 | 1 × 10⁻⁷ | rejected | 421,222 | SP 35.7 (2.2), BL 64.3 (2.2) | |
G2 | T | R10 | 0.002 | rejected | 389,918 | SP 41.2 (3.0), KAM2 58.8 (3.0) | |
G2 | T | R10 | 0.522 | fit | 333,884 | SP 36.7 (2.9), O4 63.3 (2.9) | |
G2 | T | R10 | 1 × 10⁻⁹ | rejected | 420,222 | SP 38.5 (2.4), OTT5 61.5 (2.4) | |
G2 | T | R10 | < 1 × 10⁻¹⁶ | rejected | 424,999 | SP 21.6 (3.6), Q10 78.4 (3.6) | |
G2 | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,998 | SP −639.9 (427.1), QB3 739.9 (427.1) | |
G2 | T | R10 | 4 × 10⁻⁷ | rejected | 423,551 | SP 39.9 (2.4), QI5 60.1 (2.4) |
Table 23 | Block 1, models 97–144 of 191
Cl. | T | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
G2 | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,999 | SP −18.8 (8.7), QL5 118.8 (8.7) | |
G2 | T | R10 | 0.019 | rejected | 414,857 | SP 40.1 (2.7), QP3 59.9 (2.7) | |
G2 | T | R10 | 0.060 | fit | 424,172 | SP 40.6 (2.2), QPW5 59.4 (2.2) | |
G2 | T | R10 | 5 × 10⁻¹¹ | rejected | 391,144 | SP 39.2 (3.3), QS2 60.8 (3.3) | |
G2 | T | R10 | 0.068 | fit | 407,319 | SP 41.3 (2.6), QW2 58.7 (2.6) | |
G2 | T | R10 | 3 × 10⁻⁸ | rejected | 300,024 | SP 50.6 (2.4), WAB2 49.4 (2.4) | |
G2 | T | RB13 | 9 × 10⁻¹² | rejected | 400,986 | SP 49.4 (1.6), ALA5 50.6 (1.6) | |
G2 | T | RB13 | < 1 × 10⁻¹⁶ | rejected | 405,542 | SP 51.9 (1.2), ARS16 48.1 (1.2) | |
G2 | T | RB13 | 0.864 | fit | 30,599 | SP 26.1 (6.1), AYD1 73.9 (6.1) | |
G2 | T | RB13 | 3 × 10⁻¹¹ | rejected | 405,491 | SP 39.2 (1.4), BE 60.8 (1.4) | |
G2 | T | RB13 | 3 × 10⁻¹³ | rejected | 399,227 | SP 41.9 (1.7), BL 58.1 (1.7) | |
G2 | T | RB13 | 0.006 | rejected | 373,260 | SP 43.4 (2.3), KAM2 56.6 (2.3) | |
G2 | T | RB13 | 0.005 | rejected | 307,737 | SP 40.5 (2.2), O4 59.5 (2.2) | |
G2 | T | RB13 | < 1 × 10⁻¹⁶ | rejected | 397,595 | SP 42.6 (1.7), OTT5 57.4 (1.7) | |
G2 | T | RB13 | < 1 × 10⁻¹⁶ | rejected | 405,734 | SP 36.3 (1.9), Q10 63.7 (1.9) | |
G2 | T | RB13 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 405,732 | SP −165.5 (48.1), QB3 265.5 (48.1) | |
G2 | T | RB13 | 3 × 10⁻⁸ | rejected | 404,011 | SP 44.5 (1.7), QI5 55.5 (1.7) | |
G2 | T | RB13 | < 1 × 10⁻¹⁶ | rejected | 405,734 | SP 21.4 (4.1), QL5 78.6 (4.1) | |
G2 | T | RB13 | 0.036 | rejected | 395,400 | SP 43.2 (2.0), QP3 56.8 (2.0) | |
G2 | T | RB13 | 6 × 10⁻⁴ | rejected | 404,563 | SP 45.4 (1.6), QPW5 54.6 (1.6) | |
G2 | T | RB13 | < 1 × 10⁻¹⁶ | rejected | 367,671 | SP 45.2 (2.1), QS2 54.8 (2.1) | |
G2 | T | RB13 | 5 × 10⁻¹⁰ | rejected | 381,874 | SP 47.7 (1.8), QW2 52.3 (1.8) | |
G2 | T | RB13 | 1 × 10⁻⁷ | rejected | 270,218 | SP 48.8 (1.7), WAB2 51.2 (1.7) | |
G2 | A | R10 | 2 × 10⁻⁸ | rejected | 421,441 | SP 40.0 (1.9), ALA5 60.0 (1.9) | |
G2 | A | R10 | 8 × 10⁻¹² | rejected | 424,908 | SP 40.4 (1.5), ARS16 59.6 (1.5) | |
G2 | A | R10 | 0.625 | imprecise | 31,719 | SP 27.2 (10.2), AYD1 72.8 (10.2) | |
G2 | A | R10 | 1 × 10⁻⁴ | rejected | 424,898 | SP 23.9 (1.7), BE 76.1 (1.7) | |
G2 | A | R10 | 6 × 10⁻⁶ | rejected | 421,222 | SP 25.6 (2.2), BL 74.4 (2.2) | |
G2 | A | R10 | 0.045 | rejected | 389,918 | SP 33.2 (3.1), KAM2 66.8 (3.1) | |
G2 | A | R10 | 0.254 | fit | 333,884 | SP 27.0 (3.1), O4 73.0 (3.1) | |
G2 | A | R10 | 5 × 10⁻⁸ | rejected | 420,222 | SP 28.5 (2.4), OTT5 71.5 (2.4) | |
G2 | A | R10 | 2 × 10⁻¹⁴ | rejected | 424,999 | SP 7.3 (3.2), Q10 92.7 (3.2) | |
G2 | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,998 | SP −543.5 (265.9), QB3 643.5 (265.9) | |
G2 | A | R10 | 6 × 10⁻⁵ | rejected | 423,551 | SP 29.4 (2.4), QI5 70.6 (2.4) | |
G2 | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 424,999 | SP −31.9 (8.3), QL5 131.9 (8.3) | |
G2 | A | R10 | 0.045 | rejected | 414,857 | SP 29.9 (2.7), QP3 70.1 (2.7) | |
G2 | A | R10 | 0.138 | fit | 424,172 | SP 31.4 (2.2), QPW5 68.6 (2.2) | |
G2 | A | R10 | 1 × 10⁻⁷ | rejected | 391,144 | SP 29.8 (3.4), QS2 70.2 (3.4) | |
G2 | A | R10 | 0.199 | fit | 407,319 | SP 34.0 (2.6), QW2 66.0 (2.6) | |
G2 | A | R10 | 4 × 10⁻⁹ | rejected | 300,024 | SP 41.4 (2.5), WAB2 58.6 (2.5) | |
G2 | A | RB13 | 3 × 10⁻¹⁰ | rejected | 400,986 | SP 39.6 (1.5), ALA5 60.4 (1.5) | |
G2 | A | RB13 | < 1 × 10⁻¹⁶ | rejected | 405,542 | SP 42.4 (1.1), ARS16 57.6 (1.1) | |
G2 | A | RB13 | 0.809 | fit | 30,599 | SP 17.8 (6.9), AYD1 82.2 (6.9) | |
G2 | A | RB13 | 2 × 10⁻¹² | rejected | 405,491 | SP 28.0 (1.2), BE 72.0 (1.2) | |
G2 | A | RB13 | 1 × 10⁻¹³ | rejected | 399,227 | SP 29.2 (1.7), BL 70.8 (1.7) | |
G2 | A | RB13 | 0.018 | rejected | 373,260 | SP 33.0 (2.4), KAM2 67.0 (2.4) | |
G2 | A | RB13 | 0.006 | rejected | 307,737 | SP 29.1 (2.3), O4 70.9 (2.3) | |
G2 | A | RB13 | < 1 × 10⁻¹⁶ | rejected | 397,595 | SP 31.0 (1.8), OTT5 69.0 (1.8) |
Table 24 | Block 1, models 145–191 of 191
Cl. | T | Panel | p | Result | SNPs | Weights % (SE) | Link |
|---|---|---|---|---|---|---|---|
G2 | A | RB13 | < 1 × 10⁻¹⁶ | rejected | 405,734 | SP 21.3 (1.9), Q10 78.7 (1.9) | |
G2 | A | RB13 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 405,732 | SP −164.9 (34.0), QB3 264.9 (34.0) | |
G2 | A | RB13 | 7 × 10⁻⁸ | rejected | 404,011 | SP 33.8 (1.6), QI5 66.2 (1.6) | |
G2 | A | RB13 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 405,734 | SP −0.7 (4.0), QL5 100.7 (4.0) | |
G2 | A | RB13 | 0.057 | fit | 395,400 | SP 32.4 (1.9), QP3 67.6 (1.9) | |
G2 | A | RB13 | 0.018 | rejected | 404,563 | SP 35.5 (1.5), QPW5 64.5 (1.5) | |
G2 | A | RB13 | 1 × 10⁻¹⁶ | rejected | 367,671 | SP 34.9 (2.2), QS2 65.1 (2.2) | |
G2 | A | RB13 | 5 × 10⁻⁹ | rejected | 381,874 | SP 39.0 (1.8), QW2 61.0 (1.8) | |
G2 | A | RB13 | 2 × 10⁻⁹ | rejected | 270,218 | SP 39.0 (1.7), WAB2 61.0 (1.7) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 475,229 | ALA5 −663.7 (794.8), MY 763.7 (794.8) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 478,950 | ARS16 −194.3 (58.1), MY 294.3 (58.1) | |
G2b | T | R10 | 0.058 | infeasible | 35,365 | AYD1 206.5 (102.0), MY −106.5 (102.0) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 478,950 | BE 297.5 (59.3), MY −197.5 (59.3) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 475,132 | BL 255.6 (43.9), MY −155.6 (43.9) | |
G2b | T | R10 | 4 × 10⁻¹² | rejected (infeasible) | 438,833 | KAM2 421.9 (218.8), MY −321.9 (218.8) | |
G2b | T | R10 | 1 × 10⁻⁵ | rejected (infeasible) | 377,375 | O4 268.7 (64.2), MY −168.7 (64.2) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 474,161 | OTT5 242.0 (69.7), MY −142.0 (69.7) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected | 479,024 | Q10 84.7 (6.2), MY 15.3 (6.2) | |
G2b | T | R10 | 2 × 10⁻⁵ | rejected | 479,022 | QB3 45.6 (2.2), MY 54.4 (2.2) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 477,413 | QI5 304.6 (110.0), MY −204.6 (110.0) | |
G2b | T | R10 | 5 × 10⁻¹⁰ | rejected | 479,024 | QL5 66.3 (3.3), MY 33.7 (3.3) | |
G2b | T | R10 | 3 × 10⁻¹² | rejected (infeasible) | 467,493 | QP3 340.1 (112.2), MY −240.1 (112.2) | |
G2b | T | R10 | 1 × 10⁻¹⁶ | rejected (infeasible) | 478,118 | QPW5 431.7 (168.6), MY −331.7 (168.6) | |
G2b | T | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 441,519 | QS2 205.3 (58.8), MY −105.3 (58.8) | |
G2b | T | R10 | 1 × 10⁻⁹ | rejected (infeasible) | 459,965 | QW2 433.5 (189.2), MY −333.5 (189.2) | |
G2b | T | R10 | 0.001 | rejected (infeasible) | 339,820 | WAB2 −299.8 (96.6), MY 399.8 (96.6) | |
G2b | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 475,229 | ALA5 423.5 (276.2), MY −323.5 (276.2) | |
G2b | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 478,950 | ARS16 −217.6 (113.1), MY 317.6 (113.1) | |
G2b | A | R10 | 0.141 | infeasible | 35,365 | AYD1 148.2 (58.1), MY −48.2 (58.1) | |
G2b | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 478,950 | BE 182.3 (20.5), MY −82.3 (20.5) | |
G2b | A | R10 | 9 × 10⁻¹⁴ | rejected (infeasible) | 475,132 | BL 177.3 (19.7), MY −77.3 (19.7) | |
G2b | A | R10 | 2 × 10⁻¹⁰ | rejected (infeasible) | 438,833 | KAM2 230.0 (71.8), MY −130.0 (71.8) | |
G2b | A | R10 | 4 × 10⁻⁴ | rejected (infeasible) | 377,375 | O4 198.5 (39.4), MY −98.5 (39.4) | |
G2b | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 474,161 | OTT5 149.5 (24.1), MY −49.5 (24.1) | |
G2b | A | R10 | 3 × 10⁻¹² | rejected | 479,024 | Q10 79.6 (4.2), MY 20.4 (4.2) | |
G2b | A | R10 | 2 × 10⁻⁷ | rejected | 479,022 | QB3 42.5 (2.1), MY 57.5 (2.1) | |
G2b | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 477,413 | QI5 158.1 (32.6), MY −58.1 (32.6) | |
G2b | A | R10 | 5 × 10⁻⁶ | rejected | 479,024 | QL5 63.2 (3.0), MY 36.8 (3.0) | |
G2b | A | R10 | 6 × 10⁻¹⁰ | rejected (infeasible) | 467,493 | QP3 217.1 (48.7), MY −117.1 (48.7) | |
G2b | A | R10 | 9 × 10⁻¹⁴ | rejected (infeasible) | 478,118 | QPW5 249.1 (53.9), MY −149.1 (53.9) | |
G2b | A | R10 | < 1 × 10⁻¹⁶ | rejected (infeasible) | 441,519 | QS2 121.7 (24.3), MY −21.7 (24.3) | |
G2b | A | R10 | 5 × 10⁻⁸ | rejected (infeasible) | 459,965 | QW2 268.9 (71.8), MY −168.9 (71.8) | |
G2b | A | R10 | 2 × 10⁻⁴ | rejected (infeasible) | 339,820 | WAB2 −329.5 (139.7), MY 429.5 (139.7) | |
G2c | T | R10 | < 1 × 10⁻¹⁶ | rejected | 422,516 | SP 44.4 (2.2), MY 55.6 (2.2) | |
G2c | T | RB13 | 3 × 10⁻¹¹ | rejected | 401,633 | SP 41.4 (1.7), MY 58.6 (1.7) | |
G2c | A | R10 | < 1 × 10⁻¹⁶ | rejected | 422,516 | SP 36.1 (2.4), MY 63.9 (2.4) | |
G2c | A | RB13 | < 1 × 10⁻¹⁶ | rejected | 401,633 | SP 30.4 (1.7), MY 69.6 (1.7) |
Magnus Albertus Papers · @AlbertusMvagnus · 2026
MWII0136 min read
MWII01, a low-coverage femur from Middelburg Abbey, fitted with qpAdm, f-statistics and entropy-rescaled PCA distances: a Netherlands Medieval Saxon source fits, no Slavic source adds anything, and only 4 of 104 two-source models are precise.
MAMvagnus Albertus ·
England13 min read
Three methods on the same 20 Beaker-period and 25 Middle Bronze Age persons from England: Gaussian-process match maps, entropy-rescaled PCA distances and qpAdm.
MAMvagnus Albertus ·
qpAdm6 min read
What qpAdm asks, how it computes and where it fails: a short guide to reading a qpAdm result without reading more into it than the test supports.
MAMvagnus Albertus ·
Allow analytics cookies? They help us see how the site is used so we can improve it. Your DNA and results are never included. Details