Ancestry and polygenic bias in rs-fMRI


This page summarizes the results from detecting ancestry and polygenic traits in resting state fMRI data.

In a paper published at ISBI 2019 (preprint on bioRxiv) we demonstrated that genetic ancestry, here defined as the broad four continental ancestries (Central European, East Asian, African and Native American) can be detected from resting state fMRI connectivity data. Due to an imbalance in the data, we only classified Central European ancestry vs. all other. In unpublished results we confirmed essentially the same accuracy in resting state data obtained in the UK BioBank (slides ISBI presentation).

At OHBM 2019 we extended the prediction to polygenic risk for several traits from rs-fMRI networks in healthy young adults. See the poster here.

 Posted by at 22:06