你的第一个
G25比较.

Who this guide is for

You will get the most from this manual if you:

• Want to understand what G25 coordinates are before you interpret percentages.
• Plan to use G25 Studio to compare your DNA to reference populations and run ancestry models.
• Prefer a short, practical path first, then optional detail later.

You do not need a genetics background. You do need either your own coordinates or a way to obtain them (see Before you start).

What G25 coordinates are

G25 (Global 25) coordinates are a compact numeric description of your autosomal DNA in 25 dimensions. Each dimension captures part of how your genome relates to a large reference framework built with principal component analysis (PCA).

Think of coordinates as a position in genetic space, not as a story by themselves. They are the input you feed into tools like G25 Studio. Ethnicity percentages and admixture bars are outputs of modeling your coordinates against a chosen set of reference populations. They are an interpretation layer, not the same thing as the raw coordinates.

Scaled and unscaled

• Scaled coordinates weight dimensions by how much genetic variation they explain. They are the usual choice for ancestry modeling in G25 workflows.
• Unscaled coordinates keep the raw PCA-style scale. They can be useful for some research-style comparisons.

Use coordinates that match what your calculator expects (scaled vs unscaled). Mixing types by mistake usually leads to confusing or misleading results.

What coordinates are not

• They are not a medical diagnosis.
• They are not a genealogical record or proof of a specific ancestor.
• They are not a certificate of ethnicity or nationality.

Before you start

What you need

1. A DnaGenics account and a modern web browser with JavaScript enabled.
2. G25 coordinates for the sample you want to analyze (see below).
3. A few minutes for your first run. Heavier visualizations and large population sets can take longer.

How to get G25 coordinates

Typical paths include:

• Coordinate-generation services: If you need coordinates from a supported raw DNA file, check the G25 coordinate options and Store for current availability and requirements.
• Your processed samples: If you already have coordinates from a completed service, they may appear with your sample or download area.
• Reference-style coordinates: For learning, you can use coordinates from the library inside G25 Studio (not your personal genome).

If you only have raw DNA and no coordinates yet, finish coordinate generation before expecting meaningful G25 Studio results.

Coordinate line format

G25 Studio expects one line with a name followed by 25 comma-separated numbers:

MySample_scaled,0.0234,-0.0156,0.0089,-0.0312,0.0178,0.0123,-0.0098,0.0211,-0.0145,0.0067,-0.0189,0.0256,-0.0112,0.0198,-0.0076,0.0143,-0.0221,0.0095,-0.0167,0.0134,-0.0109,0.0182,-0.0126,0.0078,-0.0094

This is a fictitious format example, not a measured result or a reference population. Copy it to inspect the format, then use your own permitted input or a reference from the library for a comparison. The important part is: one label, then 25 values, separated by commas.

Three ways to load coordinates in G25 Studio

On the main G25 Studio screen you can:

1. Paste a coordinate line into the input area.
2. Use From Library to pick curated reference coordinates.
3. Use From Samples to pull coordinates tied to your uploaded samples (when available).

These labels match the buttons in the app so you can follow along while you work.

How G25 Studio works

G25 Studio is an ancestry analysis environment: you supply coordinates, pick a calculator, adjust configuration, run the analysis, then read results.

At a high level:

Coordinates -> Calculator -> Configure -> Analyze -> Results

What is a calculator?

A calculator is a named bundle of reference populations (ancient, modern, or both) plus metadata (for example scaled vs unscaled, era, author). Your coordinates are compared to those references to produce distances, admixture-style percentages, and optional charts.

Different calculators give different results because each one only "sees" the populations it contains. A calculator focused on one world region often gives more detail there. A broad global calculator gives a wider but sometimes less granular picture. Neither is automatically "right"; they answer related questions with different tools.

Why calculator choice matters

Your genome is modeled as a mixture of the populations the calculator knows. If the calculator does not include good references for part of your ancestry, that part may be spread across other populations or show a weaker fit. That is a limitation of the reference panel, not a judgment about your DNA.

Your first beginner analysis

Follow this path the first time. You can add complexity later.

1. Load coordinates (paste, From Library, or From Samples).
2. Open Browse Calculators and pick a calculator that matches your research question. Review its references, coordinate type and scope; a broad panel can be a useful starting comparison.
3. Open the configuration area and set:
   • Algorithm: Montecarlo V1 (standard, friendly default).
   • Distance: Euclidean (good general-purpose choice).
   • Maximum populations: about 5 for easier reading.
   • Group populations: On if you want a more regional summary.
   • Ancestors Predictions (Oracle): Off for the very first run (add it when you want multi-way "what if" scenarios).
   • Analysis tools: enable PCA only for a quick visual anchor.
4. Click Calculate Results and wait until the run finishes.

This pattern mirrors the in-app Configuration Guide recommendations for a first analysis.

How to read the results

Admixture composition (Monte Carlo style)

The tool tries to reconstruct your coordinates as a blend of reference populations with flexible percentages. These are fitted weights within the chosen model, not probabilities of descent. They do not mean you have a recent ancestor from every labeled group.

Fit score

The fit score measures how well that calculator's reference set can reconstruct your coordinates. A lower residual means a closer fit within the same method, panel and settings. It does not establish a more accurate family history.

What is achievable depends on the calculator and metric. A higher residual can reflect a panel that does not represent the input well. Check the coordinate type and reference coverage, and compare alternative models. Scores from different methods or settings are not automatically comparable.

Closest populations (often linked to Oracle-style output)

Closest populations highlight references whose coordinates sit nearest to yours in G25 space. That is a neighbor idea. It is not the same as the mixture percentages, which optimize a blend. Both are useful; they answer slightly different questions.

Ancestry timeline (when shown)

If present, read a timeline alongside its method and assumptions. Historical labels or model estimates do not date a particular family connection or establish when a component entered a person's pedigree. Treat the display as research context, not a dated family tree.

Free vs PRO

Some algorithms, distance metrics, population customization, extra visualizations, and extended Oracle result counts are PRO features. If a control is locked or labeled as requiring an upgrade, you can still learn the workflow with Montecarlo V1, Euclidean distance, and PCA.

Review the Studio overview and the controls shown in your account for current capabilities. Reading this manual does not grant paid features or generate coordinates.

Monte Carlo vs Oracle

In G25 Studio, Oracle-style output is configured under Ancestors Predictions (Oracle).

• Monte Carlo (admixture-style modeling): searches for a flexible combination of populations and weights that fits your coordinates. It asks which mixture of the available references reconstructs the input under the selected settings.
• Oracle (Ancestors Predictions): explores fixed-share scenarios, such as equal splits across two, three or four reference populations. The interface label names the tool; its combinations do not identify actual parents, grandparents or other ancestors.

Use them together after you are comfortable: Monte Carlo for the main mixture read, Oracle for structured "what if" checks.

Glossary

Definitions are short on purpose. Terms appear in alphabetical order.

Admixture composition
The fitted reference-population weights used to reconstruct your coordinates within a chosen calculator.

Ancestry modeling
Using your G25 coordinates and a reference panel to estimate population-like contributions and related summaries (not the same as building a documentary family tree).

Ancestors Predictions (Oracle)
Optional analysis that scores fixed-share combinations of populations (see Monte Carlo vs Oracle).

Calculator
A named reference panel and settings bundle used to interpret your coordinates.

Closest populations
Reference populations with the smallest genetic distance to your coordinates in G25 space for the chosen method.

Distance metric
The mathematical rule used to measure how far your coordinates are from a reference population (for example Euclidean).

Euclidean distance
Straight-line distance in the 25-dimensional coordinate space. A common default.

Fit score
A summary of how well the calculator's references reconstruct your coordinates. Compare residuals within the same method, panel and settings; there is no universal range that proves genealogical accuracy.

G25 coordinates
Twenty-five numeric values (plus a sample label) describing your position in the Global25 PCA-style framework.

Genetic distance
How far apart two coordinate sets are under the chosen distance metric. Smaller usually means more similar for that metric.

Global25 (G25)
The 25-dimensional coordinate system used across G25 Studio and compatible tools.

Monte Carlo (Montecarlo)
An iterative optimization approach used to search for strong mixture fits under your chosen settings.

PCA (Principal Component Analysis)
A visualization that projects coordinates onto the main axes of variation so you can see how close you sit to reference populations.

Reference populations
Named genetic averages (ancient or modern) that define what a calculator can model.

Scaled coordinates
Eigenvalue-weighted coordinates; the usual default for G25-style modeling.

Unscaled coordinates
Coordinates on the raw PCA scale; use when your workflow intentionally calls for them.

Common beginner questions

Why do I see regions I have no family stories from?
Reference populations overlap and do not follow modern borders. Small shares can reflect shared population patterns, reference choices or model uncertainty. A small component does not establish a recent relative or a specific ancestral connection.

Why do two calculators disagree?
Each calculator uses a different reference set. Different references partition your genome differently. Compare calculators to learn robust patterns, not to pick a single "true" pie chart.

Can G25 prove my ethnicity?
No. It is a statistical comparison framework. Documentary genealogy and appropriate genetic genealogy tests serve different purposes.

Troubleshooting

• Invalid coordinate format: Confirm one name, then 25 comma-separated numbers. Remove extra blank lines or headers.
• Unexpected or unstable results: Reduce maximum populations, try group populations, and review the calculator's scope and reference coverage. Do not choose a panel solely to obtain an expected identity label.
• High fit score: Check scaled versus unscaled consistency, the metric and reference coverage. Compare alternatives within their stated limits.
• Long runtime: Disable heavy optional charts for the first pass and limit populations.

Next steps

• Review the G25 Studio overview for the wider feature context.
• Open G25 Studio to review the configuration guidance and available controls. Opening Studio does not start an analysis.
• Explore G25 reference populations.
• Browse calculator scope, authors and metadata in G25 Explorer before choosing a panel.

For more interpretation guidance, read the Ancestry FAQ. If a step is unclear, contact DNA Genics.

Disclaimer

G25 Studio is intended for education and research-style exploration. It is not a medical product. Interpret results with appropriate caution and avoid over-claiming about identity or family history based on modeled percentages alone.