Researchers worldwide can now access a free online platform that uses data from the Mexico City Prospective Study (MCPS) – the largest genetic study of a Latin American population – to accurately infer missing genetic variants from incomplete genetic datasets.
The MCPS Imputation Server uses a reference panel derived from the study to predict, or “impute”, genetic variants that were not directly measured.
Whole genome sequencing – the process of mapping an individual’s complete genetic makeup – plays an increasingly important role in identifying genetic causes of diseases and uncovering new treatment targets. However, generating and storing these data requires expensive laboratory equipment and substantial computing resources.
Instead, researchers often collect information on only a subset of genetic variants and use a process known as imputation to infer the remaining variants. Imputation servers compare partial genetic datasets with large reference panels of fully sequenced genomes. Similar to completing a puzzle by looking at the surrounding pieces, they use known patterns of genetic inheritance to accurately predict missing genetic information.
Dr Jason Torres, Senior Genetic Epidemiologist at Oxford Population Health, who leads genetic analyses of MCPS, said:
Because genetic variants are inherited together in predictable patterns, it is often possible to infer variants that have not been directly measured if you have access to a large, deeply sequenced reference dataset.
However, most existing reference panels are heavily weighted towards people of European ancestry, which means they perform less well for many other populations.
People with Latino ethnic backgrounds currently account for 2% of participants in genetic studies despite representing nearly 10% of the global population. Torres hopes the MCPS Imputation Server will advance research equity and be a valuable tool, particularly for researchers based in Central America and Mexico.
Torres added:
Some genetic causes of diseases are most likely to be identified by studying populations that have been historically underrepresented in genetic research. By providing a reference panel built from one of the world's largest studies of Mexican adults, we hope this resource will help researchers make discoveries that would otherwise be missed.
We also expect the server to be valuable for studies of diverse populations around the world, helping researchers better understand genetic risk factors that are shared across different ancestries.
The MCPS Imputation Server is the latest outcome of a long-standing collaboration between the National Autonomous University of Mexico (UNAM) and Oxford Population Health.
Established in the late 1990s, the Mexico City Prospective Study investigates the genetic and environmental causes of chronic disease in Mexico. More than 150,000 adults were recruited between 1998 and 2004.
Professor Jaime Berumen, Principal Genetic Investigator of MCPS at UNAM, said:
By making these data available through the MCPS Imputation Server, we hope to enable discoveries that would not be possible using existing reference panels alone. This resource reflects more than two decades of collaboration between researchers in Mexico and the UK and will help ensure that Mexican and other Latin American populations are better represented in the future of genomic medicine.
Development of the MCPS Imputation Server was supported by funding from the University of Oxford, Regeneron, AstraZeneca and AbbVie. Genetic sequencing was conducted by Regeneron Genetics Center. The platform is hosted by the University of Michigan and is freely available to researchers worldwide.