Localizing ancient samples in space and time from genetics with Machine Learning methods

Navn på bevillingshaver

Moisès Coll Macià


Postdoctoral Fellow


Evolutionary Biology Institute (IBE), Spanish National Research Council (CSIC) and Pompeu Fabra University (UPF).


DKK 1,100,000




Internationalisation Fellowships


The geographical location where and the time when we were born is encoded in our genomes. This is because humans tend to reproduce with close by neighbours and contemporary partners, especially in the past. In this project I aim to recover the spatial location and date of existence of an individual that lived hundreds to thousand years ago from their DNA sequences.


Nowadays we can extract and study DNA from human remains of people that lived in the past, to study our species’ evolutionary history. Unfortunately, dating and locating accurately ancient remains is not always possible, and we might then fail to contextualize the results obtained from them. The genetic method developed here will provide an alternative way to recover such valuable information.


I will develop artificial intelligence methods that will take raw DNA and output the space and time of ancient samples. This is now possible because there are many individuals for which this information is confidently known, and I can use those to train the algorithm. Thus, the method developed here will be able to learn the patterns that relate genetics to space and time.

Tilbage til oversigtssiden