Assignment 1 — Microbe Search

⬇ Rowen's .docx Saved ✓

What is actually being graded here

The assignment makes you find the same thing twice — once with AI, once with a database — and then compare. Rowen asks you to "confirm they are real articles." That is the point of the exercise: language models fabricate citations that look perfect, with plausible authors, journals, and DOIs that resolve to nothing. Open every source before you list it. The verification checklist at the bottom is the part that earns the marks.

1 · Your habitat, biome, or bioremediation process

Pick something specific enough to have a literature. "Soil" is too broad; "Antarctic Dry Valley permafrost," "human gut," "hydrothermal vent chimney," "activated sludge in wastewater treatment," or "hydrocarbon-contaminated soil undergoing bioremediation" all work. Bioremediation processes are a good choice — the papers name their organisms in the abstract.

Habitat, biome or bioremediation process selected

2 · AI-assisted search

Ask an AI model for bacteria found in your biome, and ask it to supply scientific sources — then verify what it gives you. A useful prompt shape: "List bacterial species reported in the peer-reviewed literature from [biome]. For each, give the species name and one primary research article with its DOI or PMID. Only include sources you are confident exist."

a · The prompts you used Paste the whole exchange, or at minimum every prompt. Rowen asks for this verbatim.

b · The two best sources it provided — link + full citation Confirm each one is a real article before listing it. Note here how you confirmed it (resolved the DOI, found the PMID, opened the PDF).

c · At least four bacteria found by this search

3 · Database / literature search

Now do it the traditional way — PubMed, Google Scholar, another database, or a plain browser search — and read the articles for the species names. Then compare against the AI results.

a · The method or methods you tried Name the database and give the actual query string, including any filters (date range, review vs research article, MeSH terms).

b · The two best sources found — link + full citation

c · At least four bacterial species found

d · Did both searches find the same sources and bacteria? Which was better, on speed and on credibility? A real comparison beats a diplomatic one. Speed and credibility often point in opposite directions — say so if they did for you.

Source verification checklist

Run every citation through this before you list it. Anything that fails is a fabrication or a mangled reference — replace it and say so in your write-up, because noticing is worth more than hiding it.

The DOI resolves — pasted into doi.org, it lands on the actual article.
The PMID exists in PubMed and matches the title given.
The journal is real and actually published that volume/issue in that year.
The authors match — the AI did not blend two papers into one plausible hybrid.
I opened the paper and the species I am citing is genuinely named in it — not just implied by the title.
The species names are valid — checked against LPSN. Watch for outdated genus names after the 2021 phylum reshuffle.
Species are written correctly: genus italicized and capitalized, species lowercase — Pseudomonas aeruginosa, not "pseudomonas Aeruginosa."

Where this connects to the course

This is Unit 1 material in disguise. Finding out "who lives here" without culturing anything is exactly what 16S rRNA amplicon surveys and shotgun metagenomics do — see culture-independent methods in the study guide, and 16S rRNA in Unit 2. If your biome's papers use those methods, say so; it is the kind of detail that reads as understanding rather than compliance.