The Michigan Genomics Initiative announces the release of genetic Data Freeze 7
We are excited to announce the completion of genotype calling, quality control, and genotype imputation for Freeze 7 of the Michigan Genomics Initiative dataset. This update brings significant improvements over Freeze 6, including:
An expansion of the genotyped cohort by an additional 8,442 participants, bringing the total to 88,971. The largest increases are in individuals of inferred genetic European (+6,354) and African (+969) descent, with smaller yet notable growth in East Asian (+388), Middle Eastern (+356), Central/South Asian (+254), and Native American (+121) ancestry groups.
An increase of 892,844 well-imputed genetic variants, for a new total of 53,370,970. Improvements in genotype imputation in Freeze 7 are powered by the latest R3 version of the TOPMed imputation panel. Please note that genotypes imputed in Freeze7 using the new R3 panel may exhibit allele frequency differences compared to previous freezes imputed with older panel versions. Therefore, we recommend that users do not analyze genetic data from Freeze 7 with that of older freezes.
Comprehensive genetic variant annotation for functional impact (e.g., missense, nonsense, frameshift, splice site) and clinical significance (e.g., benign, pathogenic).
For detailed cohort and technical information, please refer to the Freeze 7 white paper.
How to access this new data offering:
✅ Submit a "Custom Data Request" via JIRA: MGI Ticketing System
✅ File an IRB application via IRBMED
For assistance, contact the Research Scientific Facilitators at AIDHI-Data-Solutions@umich.edu, who can guide you through the request process.
IMPORTANT NOTE
Please be aware that as of April 24, Freeze 5 will be archived and no longer accessible through the central directory on the Armis2 compute cluster.
If you need access to Freeze 5 for essential work, please contact brettva@umich.edu, before April 24 to discuss options.
Archived data can still be requested by emailing brettva@umich.edu, but retrieval may take about a week, depending on request volume and staff availability.