In Using DNA to Determine the Origins of My Daukša/Douse Family (14 November 2026), I organized Ancestry matches from Kathleen’s, Barbara’s, and Paula’s kits into preliminary clusters at ≥25 cM and then worked through the groups most likely to reveal Eva (Douse) Rice’s maternal line or the Daukša brothers’ Lithuanian homeland. The Williams cluster tied to documented Douse cousins rather than geography; the Basgil cluster was reassigned to Peter Rice’s Račius ancestry after shared matches and the Brazavas–Bartninkai connection pointed to the Rice side; the Stramik cluster anchored a Douse-side hypothesis around Igliškėliai, Pridotkai, and nearby Daukšiai; and the Jankauskas cluster offered a working theory that Eva’s mother, recorded on the 1890 passenger list as Alistae Douse, may have been a Jankauskaitė related to Shenandoah’s Hooper and Silvenis families. At the end of that article, the Jurvic cluster remained one of three clusters in the “unresolved Rice vs. Douse” row of the summary tables—four matches on Kathleen’s kit and five on Barbara’s (up to 32 cM and 45 cM respectively)—with no side assignment and no deep dive. The present article picks up that loose end and explains how visual phasing on the X chromosome resolves it.
Why I Turned to the X Chromosome
The first article relied on AncestryDNA because that is where most of the relevant cousin kits—and the preliminary cluster assignments—live. It focused on Kathleen (my mother), her sister Barbara, and their cousin Paula. It did not discuss Kathleen’s other sister, my maternal aunt Debbie, who has also tested but not at Ancestry. All three sisters—Kathleen, Barbara, and Debbie—have their results at FamilyTreeDNA. Unlike Ancestry, FamilyTreeDNA reports the chromosomal coordinates of shared segments and includes the X chromosome in its match displays. Ancestry’s matching and clustering are based on shared autosomal DNA only; the X chromosome is not counted toward total shared cM. Two people who match each other on the X chromosome but not on the autosomes are not treated as matches at Ancestry.com at all.
While reviewing Jurvic-cluster matches on FamilyTreeDNA, I found that one member of the cluster matched both my mother and Debbie on an overlapping span of the X chromosome. That observation mattered because X-DNA in this family traces through a narrow pedigree. Margaret (Rice) Szczesny—the Lithuanian-American mother of Kathleen, Barbara, and Debbie—inherited two X chromosomes, one from each parent. She received her maternal copy from her mother, Eva (Douse) Rice, whose ancestral surname was Daukša. She received her paternal copy from her father, Peter Rice. Peter, being male, inherited a single X chromosome from his mother, Mary (ancestral surname Vilkauskas), and a Y chromosome from his father, Joseph Rice (ancestral surname Račius); fathers do not pass an X chromosome to sons.
Any X-DNA shared by Kathleen, Barbara, or Debbie with a Jurvic match must therefore come from one of two lines at Margaret’s generation: Eva’s Daukša/Douse ancestry or Mary’s Vilkauskas ancestry through Peter. It cannot represent a direct X line from Joseph Račius to Peter to Margaret, because Joseph contributed only the Y chromosome to Peter. That rule eliminates one conceivable path through Peter’s father’s side, but it does not by itself decide whether a given Jurvic segment belongs to the Daukša line or the Vilkauskas line—both remain possible until the sisters’ X chromosomes are phased and compared. The overlapping FamilyTreeDNA match nonetheless pointed to a concrete question: which of Margaret’s two X chromosomes carried the Jurvic DNA? Answering that required visual phasing among the three sisters, which is the analysis this article describes.
Visual Phasing on the X Chromosome
When parents are not available for testing, three full siblings can still reconstruct how each inherited chromosome was assembled from their father’s copy and their mother’s copy. The technique, familiar in genetic genealogy as visual phasing, works by comparing every sibling pair along one chromosome at a time and marking the boundaries where their relationship switches between half-identical and fully identical matching—each switch reflects a recombination event in at least one sibling. On the X chromosome, the logic is the same, though the pedigree constraints differ because of who can inherit an X and from whom.
I compared Kathleen, Barbara, and Debbie on the X using the chromosome browser in the Borland Genetics database, where all three kits are loaded. I treated a half-identical region (HIR) as present when shared DNA between a sister pair reached at least 7 cM; within an HIR, I called a fully identical region (FIR) only when a nested match met a 3 cM threshold. For each chromosomal interval I recorded whether a pair was HIR, FIR, or undeterminable under those rules, then summarized the results in the table below. Positions are in millions of base pairs (Mbp) on the X, using the Build 37 coordinate system.
| Start | Stop | Mom v. Barbara | Mom v. Debbie | Barbara v. Debbie |
|---|---|---|---|---|
| 2.7 | 3.5 | Undeterminable | Undeterminable | Undeterminable |
| 3.5 | 9.8 | Undeterminable | Undeterminable | Undeterminable |
| 9.8 | 33.1 | FIR | FIR | FIR |
| 33.1 | 43.6 | HIR | FIR | HIR |
| 43.6 | 88.4 | FIR | FIR | FIR |
| 88.4 | 96.5 | HIR | FIR | HIR |
| 96.5 | 97.0 | Undeterminable | Undeterminable | Undeterminable |
| 97.0 | 119.0 | FIR | HIR | HIR |
| 119.0 | 120.0 | HIR | FIR | HIR |
| 120.0 | 134.5 | FIR | HIR | HIR |
| 134.5 | 136.2 | FIR | FIR | FIR |
| 136.2 | 145.7 | HIR | HIR | FIR |
| 145.7 | 153.3 | FIR | FIR | FIR |
| 153.3 | 154.8 | HIR | HIR | FIR |
Not every interval could be called with confidence. Kathleen’s kit comes from Ancestry’s first chip version, Barbara’s from Ancestry’s second version, and Debbie’s from FamilyTreeDNA, so the three profiles do not share identical SNP coverage. Comparisons on the X are especially sensitive: some array designs undersample certain X regions, which can leave too few consecutive matching SNPs for the matcher to clear its threshold. A few intervals are also short enough that apparent agreement might reflect noise rather than inherited DNA (IBD), particularly where overlap is weak. Those stretches are marked undeterminable in red in the table. Elsewhere, almost every interval resolved to FIR or HIR.
For three full sisters, the paternal X is the same for all: each inherited Peter Rice’s single X, which came unchanged from Mary Vilkauskas. They should not display non-identical regions (NIR) with respect to that shared paternal copy—every sister pair should match on at least the paternal half, so HIR is the floor unless data quality prevents a call.
Within each determinable interval, the three sister pairs follow a strict pattern. Margaret had two distinct X chromosomes to contribute: the copy she received from Eva (Daukša/Douse ancestry) and the copy she received from Peter (ultimately Mary Vilkauskas). Each daughter’s single X chromosome is a recombination of those two templates. A fully identical region (FIR) between two sisters means they agree on both the shared Vilkauskas-derived segment from Peter and the same Eva-derived maternal mosaic across that span—both homologous sides match. A half-identical region (HIR) means they still share Peter’s Vilkauskas X in that interval, but one sister inherited Margaret’s Eva-side contribution there while the other inherited Margaret’s Peter-side contribution, so only one side matches.
With three sisters and only two grandparental X sources at Margaret’s generation, each interval should show either three FIR pairings (all three sisters inherited the same maternal mosaic across the span) or one FIR and two HIR pairings (two sisters match on the maternal mosaic and the third received the alternate copy). You cannot see HIR for all three pairs in the same interval: that would require three different maternal X outcomes where only two templates exist.
Maternal X Matches
With the sister chart in hand, I next lined up X-chromosome matches that could be assigned to the maternal half of each sister’s DNA. For Kathleen and Barbara, I relied chiefly on cousins Ancestry SideView had already classified as maternal, supplemented by close relatives whose family connections I know well enough to treat as maternal for this purpose. Debbie’s list comes from FamilyTreeDNA coordinates for the same individuals where she matches on the X. The table lists each match once per sister with the overlapping span on the X in millions of base pairs (Build 37).
| Mom (Kathleen) | Barbara | Debbie |
|---|---|---|
|
|
|
S.H. appears in the Jurvic cluster that prompted this study. On FamilyTreeDNA she overlaps Kathleen and Debbie on the X within the spans shown, which is what first drew my attention to X-based phasing for this cluster.
Laura requires a footnote from the earlier investigation. In the preliminary assignment tables for the first article, I counted her among Kathleen’s Eva Douse singletons—matches at ≥25 cM with no cluster mates, filed under unresolved John Douse vs. Alistė and suspected as a possible Daukša-side cousin. That placement was never confirmed through a full clustering pass. On Ancestry I had looked only at her overlap with my mother and with Barbara on chromosome 7; neither sister had other strong shared matches across that same segment, so Laura never appeared as part of a cluster on either kit.
Assigning Vilkauskas vs. Daukša on the Maternal X
The anchor match is Nancy, a documented descendant of Peter Rice’s sister Mary Rice. That Mary Rice is not Peter’s mother, Mary (Vilkauskas) Rice—the woman who passed her X chromosome to Peter and who is the subject of the reconstruction below. Nancy does not descend from Eva (Douse) Rice or from John Douse’s family. Setting aside a hypothetical close kinship between the Vilkauskas and Daukša lines in Lithuania—which I do not consider likely for several independent reasons—any X-DNA overlap between Nancy and one of the three sisters tells us that the matching interval on the sister’s maternal X is Vilkauskas-associated (Peter’s side, through Mary Vilkauskas), not Daukša-associated through Eva.
Nancy’s intervals bracket most of the maternal X where the other cousins also match. Leona’s and Adele’s spans fall inside the same Vilkauskas-colored regions once Nancy’s boundaries are applied, so I treat their X matches the same way. S.H. and the other SideView-maternal matches listed for Debbie overlap only intervals that the phasing chart assigns to Vilkauskas; none of the maternal cousins in the table land on a sister’s Daukša blocks.
With three full sisters, the FIR/HIR worksheet left relatively little ambiguity: recombination points are sparse on the X, Nancy’s overlaps pin whole segments to Vilkauskas, and the two-sister-versus-one pattern across each interval narrows who received Margaret’s Eva-side copy versus her Peter-side copy in the remaining spans. The table below shows the result for every determinable interval—blue for Vilkauskas (Peter’s maternal line, blocks 1 and 3), yellow for Daukša (Eva’s line, blocks 2 and 4), red where the sister comparison could not be called.
| Start | Stop | Mom | Barbara | Debbie |
|---|---|---|---|---|
| 2.7 | 3.5 | Undeterminable | Undeterminable | Undeterminable |
| 3.5 | 9.8 | Undeterminable | Undeterminable | Undeterminable |
| 9.8 | 33.1 | Block 1 — Vilkauskas | Block 1 — Vilkauskas | Block 1 — Vilkauskas |
| 33.1 | 43.6 | Block 1 — Vilkauskas | Block 2 — Daukša | Block 1 — Vilkauskas |
| 43.6 | 88.4 | Block 1 — Vilkauskas | Block 1 — Vilkauskas | Block 1 — Vilkauskas |
| 88.4 | 96.5 | Undeterminable | Undeterminable | Undeterminable |
| 96.5 | 97.0 | Undeterminable | Undeterminable | Undeterminable |
| 97.0 | 119.0 | Block 3 — Vilkauskas | Block 3 — Vilkauskas | Block 4 — Daukša |
| 119.0 | 120.0 | Block 4 — Daukša | Block 3 — Vilkauskas | Block 4 — Daukša |
| 120.0 | 134.5 | Block 3 — Vilkauskas | Block 3 — Vilkauskas | Block 4 — Daukša |
| 134.5 | 136.2 | Block 3 — Vilkauskas | Block 3 — Vilkauskas | Block 3 — Vilkauskas |
| 136.2 | 145.7 | Block 4 — Daukša | Block 3 — Vilkauskas | Block 3 — Vilkauskas |
| 145.7 | 153.3 | Block 3 — Vilkauskas | Block 3 — Vilkauskas | Block 3 — Vilkauskas |
| 153.3 | 154.8 | Block 4 — Daukša | Block 3 — Vilkauskas | Block 3 — Vilkauskas |
Every maternal cousin in the match table—including Laura and S.H.—falls on Vilkauskas intervals for the sister who matches her. Laura’s tentative Daukša-side classification in the November article was therefore wrong; I updated my chromosome-7 maps for her accordingly once the X placement was clear.
Conclusion: the Jurvic Cluster Is Rice / Vilkauskas
Based on Nancy’s documented relationship, the shared overlap pattern among Leona, Adele, and S.H., and the phased maternal X, the evidence supports a single assignment: members of the Jurvic cluster are related to Margaret (Rice) Szczesny and her three daughters through Peter Rice’s Vilkauskas line, not through Eva’s Daukša/Douse ancestry. That removes one of the three “unresolved Rice vs. Douse” clusters from the first article’s inventory and reclassifies it on the Rice side of the house—specifically with Mary Vilkauskas’s kin.
Reconstructing Mary (Vilkauskas) Rice
As a last step I used Borland Genetics reconstruction tools to export the Vilkauskas-side segments from the sisters’ chromosome maps into one synthetic kit representing Margaret’s paternal grandmother Mary (Vilkauskas) Rice—the woman who passed her X to Peter and, through him, passed blocks of that genetic material to her granddaughters. Genome coverage in the reconstruction is about 4%.
| Chromosome | Start | Stop |
|---|---|---|
| 1 | 87,746,316 | 98,808,594 |
| 2 | 51,554,078 | 63,126,755 |
| 2 | 126,572,825 | 153,767,697 |
| 4 | 6,558,428 | 31,709,061 |
| 6 | 148,303,942 | 157,444,915 |
| 8 | 21,488,272 | 40,569,598 |
| 9 | 23,182,825 | 30,613,670 |
| 13 | 72,565,456 | 111,110,817 |
| 23 (X) | 9,757,295 | 88,383,216 |
| 23 (X) | 96,422,059 | 154,854,340 |
Taken together, those intervals total about 325 cM, including roughly 150 cM on the X chromosome across the two spans listed above. Among the three sisters, Barbara shares the most with the reconstructed kit: 254 cM overall, 108 cM of it on the X alone. I match the reconstruction at 77 cM, including 46 cM on my X—but not the interval that drives male-pattern baldness. I have a different maternal ancestor to blame for that!