AfriGen-D FedImpute

Federated Genotype
Imputation Platform

Discover reference panels, access data, and run pipelines across federated services -- built on GA4GH and community standards.

DRS v1.5.0WES v1.1.0Service InfoData Connect v1.0.0TRS v2RO-Crate
What you can run
  • Genotype ImputationAvailable now
  • HLA Type ImputationComing soon
  • PRS CalculationComing soon

Where do you want your data imputed?

One account, one job wizard, one place to track results. You choose the service for each job.

African-hosted

AfriGen-D imputation service

Your data stays on African infrastructure, and the African reference panels never leave it.

Hosted at
ILIFU research cloud, University of Cape Town
Panels
H3Africa v6 and v7, 1000 Genomes NYGC 30x
You need
Your AfriGen-D sign-in. Nothing else.
Cost
Free for academic researchers
US-hosted

Michigan Imputation Server

Submit from FedImpute to the Michigan Imputation Server and use the panels it offers.

Hosted at
The server's own US infrastructure
Panels
HRC r1.1, 1000 Genomes, HapMap 2, CAAPA and more
You need
Your own Michigan account token
Your data
Sent to this server after you confirm the transfer. Its terms apply.
US-hosted

TOPMed Imputation Server

Submit from FedImpute to the TOPMed Imputation Server and use the TOPMed reference panel.

Hosted at
The server's own US infrastructure
Panels
TOPMed r3
You need
Your own TOPMed account token
Your data
Sent to this server after you confirm the transfer. Its terms apply.

The job wizard shows where each service is hosted and asks you to accept the transfer terms before anything leaves AfriGen-D. Jobs on every service appear in the same list.

How it works

Three GA4GH standard APIs, one seamless workflow

1

Discover

Data Connect

Query reference panels by build, population, and continent across federated services.

2

Access

DRS + Passport

Authenticate via SSO. Resolve DRS URIs for panels and input files with visa-based access control.

3

Compute

WES + RO-Crate

Submit imputation jobs to federated services. Track progress. Download results with full provenance.

Quick Start Guide

Get started with genotype imputation in 4 simple steps

1
Create Account
Sign up for a free account. You'll need to verify your email address before you can submit jobs.
2
Prepare Your Data
Format your genotype data as VCF files. Ensure coordinates match hg19 or hg38. Use our Checkref tool to validate.
3
Submit Job
Upload your data once, then choose where to run it -- add more than one service and reference panel to the same submission if you need them.
4
Download Results
Receive email notification when complete (1-3 hours). Download imputed genotypes and QC reports using your password.

Need help? Check out our detailed documentation or contact support.

Everything you need for genomic imputation

A federated platform connecting African compute infrastructure via GA4GH and community standards.

Federated Services
Run imputation across African services, each exposing WES, DRS, and Data Connect endpoints, or send a job to the Michigan and TOPMed servers.
GA4GH Native
Built on 5 GA4GH standards (DRS, WES, Service Info, Data Connect, TRS) and RO-Crate.
Panel Discovery
Query reference panels across all services via Data Connect. Filter by build, population, and continent.
African Reference Panels
H3Africa v7 on GRCh38: 4,283 samples and 69.6M variants, plus a 100% African v6 subset spanning 22 countries. Purpose-built for African genomics.
Real-time Monitoring
Track job progress with live status from WES. RO-Crate provenance captures every step.
Visa-Based Access
GA4GH Passport/Visa controls access to controlled datasets. Open data for all, restricted data by visa.

African-hosted, no commercial cloud dependency

The AfriGen-D service runs on ILIFU research cloud at the University of Cape Town. On it, data stays in South Africa under POPIA. Free for academic researchers.

Supported by NIH grant U24HG012750

Provided to you by

AfriGen-D
Interconnected resources and services for African genomics research data lifecycle management and implementation.
H3ABioNet
Pan African Bioinformatics Network to advance bioinformatics research and capacity building in Africa.
H3Africa
Consortium of 48 African projects aiming to enhance the health of African populations.
ilifu
Cloud computing research infrastructure for Astroinformatics and Bioinformatics hosted at UCT ICTS, managed by IDIA.

Frequently asked questions

Who can use AfriGen-D FedImpute?+

Academic and non-profit researchers studying human genomics can register for free. Each controlled-access reference panel has its own Data Access Agreement (DAA) that must be accepted in-app before use.

Which reference panels are available?+

On the AfriGen-D service: H3Africa v6 (full and African-only subsets) and H3Africa v7 for African-ancestry imputation, and 1000 Genomes NYGC 30x. The Michigan and TOPMed servers add the panels they offer, such as HRC r1.1, the 1000 Genomes panels, HapMap 2, CAAPA and TOPMed r3. More H3Africa panels are added as the H3Africa Consortium publishes them.

Where is my data stored?+

Uploaded VCFs, imputation results, and provenance records are stored on the ILIFU research cloud at the University of Cape Town, in compliance with POPIA and H3Africa governance. On the AfriGen-D service, reference panels never leave African infrastructure. If you send a job to the Michigan or TOPMed server instead, your data goes to that US-hosted server after you confirm the transfer. You can download your own imputed results (encrypted with a one-time password) while they are available on the platform.

How long are my results kept?+

Input files are retained for 30 days after upload. Imputed results are retained for 7 days after completion. Results are encrypted with a one-time password delivered via email -- download them before expiry.

What file formats are supported?+

VCF or bgzipped VCF (.vcf.gz). Files must be sorted by chromosomal position, use a supported reference build (hg19 or hg38), and contain at least 20 samples.

Is there a cost?+

No. AfriGen-D FedImpute is free for academic researchers, supported by NIH grant U24HG012750.

Can I also use the Michigan or TOPMed imputation servers?+

Yes. Add your own account token for that server in Settings, pick its panel in the job wizard, and confirm the data transfer. Your data is then sent to that US-hosted server and its terms apply. FedImpute itself is built on open GA4GH standards (DRS, WES, Passport, Data Connect), so one submission can run across several services and panels, each with its own settings. Only the AfriGen-D service keeps your data on African infrastructure.