See what your DNA can actually tell you
Nine things we can work out from a data file. Every one of them has a free demo on this page that you can play with right now, with no sign-up and no upload. Scroll down and try them.
Which question is yours?
Pick the one that sounds like what you want to know. Each card jumps to a demo you can play with.
- 01
Where did this person's ancestors actually come from?
Ancestry Modeling (qpAdm & f-statistics)
Try the demo - 02
When did two peoples become one?
Admixture Dating
Try the demo - 03
Which fatherline and motherline is this?
Y-DNA & mtDNA Haplogroup Analysis
Try the demo - 04
The DNA is damaged. Can anything still be done with it?
Low-Coverage Imputation & Rescue
Try the demo - 05
Were these people family?
Kinship & Relatedness
Try the demo - 06
Where do I fit in the ancient world?
Personal Ancient Ancestry Reports
Try the demo - 07
Does this molecule actually stick to its target?
Molecular Dynamics (MD Simulations)
Try the demo - 08
What shape does this molecule take, and what does it cost?
Density Functional Theory (DFT)
Try the demo - 09
How will this material behave before we build it?
Materials Simulations
Try the demo
Where did this person's ancestors actually come from?
Ancestry Modeling (qpAdm & f-statistics)
We take a genome and work out which ancient peoples mixed together to produce it, and in what proportions. The part that matters is that every answer is put through a statistical test, so you learn not only what fits the data but what the data rules out.
Step 1
You send a DNA file
Plus the question you actually want answered.
Step 2
We test many recipes
Hundreds of candidate mixtures, each accepted or rejected.
Step 3
You get ranked models
With proportions, error margins, and every rejected rival listed.
Try it, then read a study we actually ran
Load the demo file to see what we accept and what we read out of it. Then run it, and you get the real report from this service, not a mock-up.
- 1Pick a file
- 2Run it
- 3See your results
EIGENSTRAT genotypes
The format published ancient-DNA datasets ship in, including the Allen Ancient DNA Resource. Deliberately full of gaps, exactly like real ancient data.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
A worked demonstration on public reference data. It is not a record of any client, and the population names refer to archaeological samples, not to any modern nation or person.
Best for: researchers preparing publications · labs without a population-genetics specialist · advanced genealogists with deep questions
Technical detail, for specialists
File formats accepted
- BAM / CRAM
- VCF
- EIGENSTRAT (.geno/.snp/.ind)
- PLINK (.bed/.bim/.fam)
- 23andMe / FamilyTreeDNA raw export
Software we run
- qpAdm
- qpWave
- f3 / f4 / D
- ADMIXTOOLS 2
- PCA
When did two peoples become one?
Admixture Dating
If you already know that two groups mixed, this works out when it happened. The DNA carries its own clock: the shared stretches of chromosome get chopped shorter with every generation, and the speed of that shortening gives you a date.
Step 1
You send genomes
The mixed population, and the two source groups.
Step 2
We measure the decay
How fast the shared blocks have broken up.
Step 3
You get a date range
With a confidence interval, read against the archaeology.
Try it, then read a study we actually ran
Load the demo file to see what we accept and what we read out of it. Then run it, and you get the real report from this service, not a mock-up.
- 1Pick a file
- 2Run it
- 3See your results
EIGENSTRAT genotypes
The format published ancient-DNA datasets ship in, including the Allen Ancient DNA Resource. Deliberately full of gaps, exactly like real ancient data.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
A worked demonstration on public reference data. It is not a record of any client, and the population names refer to archaeological samples, not to any modern nation or person.
Best for: studies linking genetic mixture to historical events: migrations, conquests, trade-network eras
Technical detail, for specialists
File formats accepted
- EIGENSTRAT / PLINK genotypes
- VCF
- Phased data welcome, not required
Software we run
- DATES
- ALDER-class LD decay
Which fatherline and motherline is this?
Y-DNA & mtDNA Haplogroup Analysis
Your direct fatherline and your direct motherline each carry their own unbroken record, passed down untouched for thousands of years. We read both and place them precisely on the human family tree, showing the evidence behind every branch rather than just handing you a label.
Step 1
You send a DNA file
Ancient or modern, damaged or clean.
Step 2
We test the markers
Several independent tools, cross-checked against each other.
Step 3
You get a placement
The branch, the evidence at each step, and what it means.
Try it, then read a study we actually ran
Load the demo file to see what we accept and what we read out of it. Then run it, and you get the real report from this service, not a mock-up.
- 1Pick a file
- 2Run it
- 3See your results
Consumer DNA export
What you download from 23andMe, AncestryDNA, MyHeritage, FamilyTreeDNA or Living DNA. If you have taken a home DNA test, your file looks like this one.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
A worked demonstration on public reference data. It is not a record of any client, and the population names refer to archaeological samples, not to any modern nation or person.
Best for: surname/tribal lineage projects · aDNA studies needing uniparental calls · genealogists past the limits of consumer reports
Technical detail, for specialists
File formats accepted
- BAM / FASTQ
- Big-Y / WGS results
- VCF
Software we run
- Yleaf
- pathPhynder
- yHaplo
- HaploGrep
- damage-aware calling
The DNA is damaged. Can anything still be done with it?
Low-Coverage Imputation & Rescue
Ancient and degraded samples come back full of holes. We fill those holes by comparing what survived against thousands of complete genomes, then tell you plainly which analyses the rescued data can support and which it cannot.
Step 1
You send thin data
Shallow shotgun, degraded, or budget-limited sequencing.
Step 2
We fill the gaps
Using large reference panels, with quality scored per position.
Step 3
You get a usable genome
Plus an honest statement of its limits.
See the shape of the report
A worked example for this service is in preparation. Until then, this shows how the finished report is laid out.
- 1Pick a file
- 2Run it
- 3See your results
EIGENSTRAT genotypes
The format published ancient-DNA datasets ship in, including the Allen Ancient DNA Resource. Deliberately full of gaps, exactly like real ancient data.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
The figures above are a worked example from our synthetic demo file, shown so you can see the shape of a finished report. A real job runs on a compute cluster against reference datasets measured in gigabytes, and takes days. Nothing here is a result about a real person or sample.
Best for: labs with shallow-shotgun ancient samples · degraded or budget-limited sequencing projects
Technical detail, for specialists
File formats accepted
- BAM / CRAM (usable from ≥0.1×)
- FASTQ
Software we run
- GLIMPSE2
- 1000G / HRC panels
- validation vs. direct calls
Were these people family?
Kinship & Relatedness
Given two or more genomes, we measure how much DNA they share and turn that into an actual relationship: parent and child, siblings, cousins, or no relation at all. Then we work out which family trees are consistent with every pair at once.
Step 1
You send the genomes
Any number of individuals from a site or a family.
Step 2
We compare every pair
Shared DNA measured, degree of relationship assigned.
Step 3
You get a family tree
Every pairing, plus the pedigrees that fit them all.
See the shape of the report
A worked example for this service is in preparation. Until then, this shows how the finished report is laid out.
- 1Pick a file
- 2Run it
- 3See your results
VCF variant calls
The standard research format, holding three individuals. This is what a sequencing provider or a bioinformatics core usually hands back.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
The figures above are a worked example from our synthetic demo file, shown so you can see the shape of a finished report. A real job runs on a compute cluster against reference datasets measured in gigabytes, and takes days. Nothing here is a result about a real person or sample.
Best for: cemetery & burial studies · family-structure questions · verification of suspected relationships
Technical detail, for specialists
File formats accepted
- BAM per individual
- VCF
- EIGENSTRAT / PLINK genotypes
Software we run
- READ
- KIN
- lcMLkin
- IBD methods
Where do I fit in the ancient world?
Personal Ancient Ancestry Reports
Upload the raw file from a consumer DNA test and we model your ancestry against real ancient genomes, using the same methods a published study would use. What comes back is an illustrated report with maps, dates and citations, not a coloured pie chart.
Step 1
You upload raw data
From 23andMe, FamilyTreeDNA, Ancestry, or a sequencing provider.
Step 2
We model it properly
A formal, tested ancestry model against ancient references.
Step 3
You get an illustrated report
Maps, context, and the published papers behind each claim.
See the shape of the report
A worked example for this service is in preparation. Until then, this shows how the finished report is laid out.
- 1Pick a file
- 2Run it
- 3See your results
Consumer DNA export
What you download from 23andMe, AncestryDNA, MyHeritage, FamilyTreeDNA or Living DNA. If you have taken a home DNA test, your file looks like this one.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
The figures above are a worked example from our synthetic demo file, shown so you can see the shape of a finished report. A real job runs on a compute cluster against reference datasets measured in gigabytes, and takes days. Nothing here is a result about a real person or sample.
Best for: individuals & families · tribal/regional heritage projects · anyone whose curiosity outgrew the pie chart
Technical detail, for specialists
File formats accepted
- 23andMe raw data (.txt/.zip)
- AncestryDNA raw data
- MyHeritage / FamilyTreeDNA export
- Living DNA / TellMeGen export
Software we run
- 23andMe / FamilyTreeDNA / Ancestry uploads
- qpAdm
- PCA vs. ancient refs
Does this molecule actually stick to its target?
Molecular Dynamics (MD Simulations)
A drug only works if it stays attached to the protein it is aimed at. We simulate every atom moving, at real body temperature, across a stretch of time, and watch whether it holds on or falls off. It is a way to test the idea before anyone spends money in a laboratory.
Step 1
You send structures
A protein, and the molecule you want tested against it.
Step 2
We run the movie
Every atom simulated, then measured for stability and binding.
Step 3
You get the trajectory
With stability analysis and full logs for reproducibility.
See the shape of the report
A worked example for this service is in preparation. Until then, this shows how the finished report is laid out.
- 1Pick a file
- 2Run it
- 3See your results
Protein with a ligand
A protein structure with a small molecule docked beside it. This is what a binding simulation starts from.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
The figures above are a worked example from our synthetic demo file, shown so you can see the shape of a finished report. A real job runs on a compute cluster against reference datasets measured in gigabytes, and takes days. Nothing here is a result about a real person or sample.
Best for: structural biologists · drug-discovery and biotech teams · researchers studying mutation effects or ligand binding
Technical detail, for specialists
File formats accepted
- PDB / mmCIF structure
- Ligand file (SDF / MOL2)
- FASTA sequence, if we build the model
Software we run
- GROMACS
- AMBER
- NAMD
- OpenMM
- Protein-Protein
- Protein-Ligand
- Protein-Peptide
- Protein-DNA/RNA
- Lipid Bilayer
What shape does this molecule take, and what does it cost?
Density Functional Theory (DFT)
Quantum chemistry worked out from first principles: the resting shape a molecule settles into, what its electrons are doing, and how much energy a reaction takes or releases. No experiment required, and the maths is the same maths used in published chemistry.
Step 1
You send a structure
Or just a description of the chemistry in question.
Step 2
We solve it
Geometry optimised, electronic structure and energetics calculated.
Step 3
You get energies
Optimised coordinates, plus every input and output file.
See the shape of the report
A worked example for this service is in preparation. Until then, this shows how the finished report is laid out.
- 1Pick a file
- 2Run it
- 3See your results
Molecule coordinates
A single molecule as plain atom coordinates. The usual starting point for a quantum chemistry calculation.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
The figures above are a worked example from our synthetic demo file, shown so you can see the shape of a finished report. A real job runs on a compute cluster against reference datasets measured in gigabytes, and takes days. Nothing here is a result about a real person or sample.
Best for: computational chemists · materials researchers · groups needing a second opinion on a DFT result
Technical detail, for specialists
File formats accepted
- XYZ / molecular structure
- CIF for periodic systems
- Or simply describe the system
Software we run
- ORCA
- Quantum ESPRESSO
- Geometry optimization
- Electronic structure
- Reaction energetics
How will this material behave before we build it?
Materials Simulations
We build a material atom by atom, add the kind of flaws that real manufacturing produces, and test its strength, stiffness and heat behaviour in simulation. You find out how it fails before a physical sample exists.
Step 1
You send a structure
A crystal, an alloy, or the property you need predicted.
Step 2
We simulate it
Defects, strain, and thermal behaviour, at atomic scale.
Step 3
You get property values
With uncertainty, full logs, and methods text for publication.
See the shape of the report
A worked example for this service is in preparation. Until then, this shows how the finished report is laid out.
- 1Pick a file
- 2Run it
- 3See your results
Crystal structure
A repeating crystal lattice with its unit cell and symmetry. What a materials calculation is built on.
Have your own data? You do not need it here. Send it to us and we will run the real analysis on it.
The figures above are a worked example from our synthetic demo file, shown so you can see the shape of a finished report. A real job runs on a compute cluster against reference datasets measured in gigabytes, and takes days. Nothing here is a result about a real person or sample.
Best for: materials science researchers · R&D teams evaluating a new material · groups needing simulation support alongside experiment
Technical detail, for specialists
File formats accepted
- CIF
- POSCAR
- XYZ / structure description
Software we run
- LAMMPS
- Quantum ESPRESSO
- ASE
- Crystal structures & defects
- Mechanical & thermal properties
How a project runs
- 1Day 0
Request
Open a project in the Client Portal: goals, samples, data type. Free.
- 2≤ 48 h
Scoped quote
Fixed price and timeline for exactly the analyses agreed. No payment yet.
- 3Days 2–12
Analysis
You chat directly with your analyst; interim findings land in your thread.
- 4Delivery
Report + logs
PDF report, results workbook, full command logs, methods text. Revisions included.
Common questions
What data formats do you accept?
FASTQ, BAM/CRAM, VCF, EIGENSTRAT/PLINK datasets, and raw exports from 23andMe, AncestryDNA, MyHeritage, FamilyTreeDNA, Living DNA, and TellMeGen. Not sure what you have? Upload it and we'll triage it for free.
Is my genetic data safe with you?
Uploads go into a private, encrypted storage space visible only to you and your assigned analyst. Never public, never shared, never used beyond your project. You can request deletion of all your files at any time, and we confirm in writing when it's done.
How is pricing determined?
By scope: number of samples, data type, and analyses requested. You receive a fixed quote before any work starts, and the price doesn't move unless you change the scope.
Can I use your results in a publication?
Yes, that is the point. Deliverables include methods text with citations, full command logs, and parameter files, so your methods section and your reviewers are both covered. We can also join as co-authors or acknowledged analysts, whichever fits your policy.
My samples are very low coverage. Is anything possible?
Usually, yes. Imputation (Service 04) rescues surprisingly shallow data, and some methods (haplogroup calling, READ kinship, f-statistics on transversions) are built for exactly this situation. Send us the coverage numbers and we'll tell you honestly what's feasible.
Do you handle molecular dynamics / protein simulation work too?
Yes. Alongside genomics we run molecular dynamics simulations in GROMACS, AMBER, NAMD, or OpenMM (protein stability, mutation comparison, ligand studies). Contact us with your system and we'll scope it like any other project.
Do you also run DFT calculations?
Yes. We run density-functional-theory calculations for electronic structure, geometry optimization, and reaction energetics, using standard plane-wave and localized-basis codes. Send us your system and the properties you need, and we'll scope compute time and turnaround like any other project.
What about materials simulations?
Yes. Beyond biomolecules, we run classical and ab-initio materials simulations, crystal structures, defects, mechanical and thermal properties, on the same infrastructure we use for our genomics and MD work. Tell us the material and the property you're after and we'll scope it.
Not sure which one fits? Just describe your question.
Write to us in plain language. We will map it to the right analyses and quote it within 48 hours.
Start a project