Post

Happy to announce that MEDI, our method to quantify dietary intake through metagenomics is out now. 🧬 🧪🧫🦠 www.nature.com/articles/s42... #metagenomics #microbiome #microbiomesky So what is MEDI? A combination of two things:

nature.comMetagenomic estimation of dietary intake from human stool - Nature MetabolismDiener et al. present a method that allows the estimation of dietary intake from human stool by detecting food-derived DNA in faecal metagenomes.

(1) a linked food database of more than 400 food items that includes nutrition and genomic information with multiple fallback strategies (partial assemblies, higher ranks)

(2) a high specificity identification and quantification pipeline that reliably identifies metagenomic reads coming from food items even when those make up less than 0.001% of the total DNA

0 reposts2 likes

One new things we added since the preprint is validation with two controlled feeding studies. In both, MEDI was able to detect the dietary intervention . There was also good agreement with some micro- and macronutrients when creating a standardized portion out of the detected food items.

We were also able to infer the onset of solid food consumption in infants and found good agreement of MEDI with food frequency questionnaires. Last but not least, MEDI could identify dietary patterns associated with metabolic syndrome in a retrospective study.

Of course, there are limitations too. We can't detect foods or additives without DNA such as added sugars or fats for instance. Also there is quite some adsorption and breakdown in the intestinal tract. Building the database is also a bit of a drag and we are looking for ways to make this easier.

Right now MEDI works best with high depth sequencing (>30M reads) and only works for human samples (working on including other hosts). However, it does *not* require new assays and can be used for existing metagenomic sequencing data to add dietary information.