A new method for analyzing petrified feces will help to accurately determine its source – in other words, to understand to whom it belonged to a person or a certain type of animal. The development is described in a study by scientists from the Max Planck Institute of Natural Sciences, published in the journal BIOPHYSICS AND MOLECULAR BIOLOGY.
Archaeologists during excavations often find petrified feces. Such findings are of great value for science – their analysis can show how certain species of animals or human ancestors ate. In addition, feces can restore the evolution of parasites.
So far, the main problem for the use of these data in scientific research has been the lack of a method that would allow you to accurately determine to whom they belonged.
In a new study, scientists found a solution to this problem. They developed the coproID platform, which, using machine learning, allows you to analyze the host DNA and compare the data with the genome of living organisms known to science.
The AI, which underlies the system, was trained on the microbiome data of modern dog fecal particles. Testing the method showed that data on the combination of host DNA and various colonies of microbes living in the intestines of humans and dogs can accurately distinguish their excrement.
“Thanks to additional data on the intestinal metagenomes of non-western rural dogs, we can better classify even older canine feces is actually canine feces, and not as“ indeterminate ones ”. As the data on the microbiomes of other creatures grows, we can improve coproID and help researchers who encounter pale phases in different geographic and historical contexts”.
Maxim Borri, lead author of the study