FoodSeq Handbook
FoodSeq identifies the plants and animals in a person's diet from DNA in stool. This handbook documents the full workflow, from preparing sequencing libraries to an analysis-ready table of foods, so that anyone can generate, process, and compare dietary metabarcoding data. All protocols, code, and reference data are open source.
The instructions assume minimal familiarity with R and the shell. This site is actively maintained by Edible Atlas at Duke University, and pages are updated as methods develop.
What's here
- Library preparation protocols for two markers: trnL (plants) and 12SV5 (vertebrates). See Wet Lab Protocols.
- A pipeline that takes sequencing output to a phyloseq object, plus a tool that removes human reads before public deposition: mb-pipeline. See Creating a Phyloseq.
- A curated reference database of human foods for both markers: food-dbs, with versioned releases archived on Zenodo. See Creating the References.
- A common-name map that turns sequences into recognizable food names. See Assigning Common Names.
- An R package, foodseq.tools, for quality control, taxonomy assignment, common names, harmonizing related sequences across runs, and PCA. See Post-Phyloseq Processing and Analysis.
Where to start
- Starting from samples: Wet Lab Protocols, then Creating a Phyloseq, then Post-Phyloseq Processing and Analysis.
- Starting from sequencing data: Creating a Phyloseq, then Post-Phyloseq Processing and Analysis.
- New to DNA metabarcoding, R, GitHub, or a computing cluster: start with Getting Started.
General and Duke instructions
Some pages have two tabs. General works at any institution; Duke shows the setup used at Duke.
Citing
If you use the tools or protocols described here, please cite this handbook using the DOI above. If you use the reference database, please also cite its archived release (10.5281/zenodo.22756737).
Feedback and contributions
If you have comments, questions, or suggestions, use the widget at the bottom of any page or open an issue on the GitHub repository. To suggest foods or sequences for the reference database, open an issue on food-dbs.