Welcome to the MecCogChallenge!

About This Challenge

Welcome to the MecCog Challenge, a collaborative effort to map the mechanisms by which the APOE4 genetic variant increases the risk of Alzheimer's disease.

What You'll Do

Your task is to find relevant research papers and extract key experimental findings that relate to specific APOE4-Alzheimer's hypothesis you decide to select.

For each hypothesis, you will:

  1. Search the scientific literature (PubMed, preprints, databases, web sources)
  2. Identify papers that bear on your specific mechanism
  3. Extract experimental findings from those papers (e.g., measurements, effects, experimental conditions)
  4. Record everything in a structured spreadsheet that we'll validate

This is Step 1 of a multi-phase challenge. Later phases will ask you to assess evidence, estimate mechanism confidence, and design experiments — but for now, we focus on finding and documenting what is already known.

The Hypotheses

Hypotheses will be released in batches starting September 2026. Check for new communications.

Focus on the hypothesis you decide to work on. Don't catalog Alzheimer's mechanisms in general — we need findings that directly test or inform the specific hypothesis you picked.

Challenge Rules

What You Must Do

✓ Search thoroughly — use PubMed, preprint servers (bioRxiv/medRxiv), Google Scholar, databases, and any sources you think are relevant.

✓ Extract precise findings — include:

  • A one-sentence description of each finding
  • A direct quote from the paper (or mark N/A if not applicable)
  • The experimental system tested (e.g., "primary mouse microglia", "human iPSC-derived neurons")
  • Where in the paper the finding appears (figure panel, table, or text excerpt)
  • Statistical details if available (effect size, P value, sample size)

✓ Fill the submission spreadsheet — use the official template provided. Every field must have a value or the literal N/A (never leave blanks).

✓ Record metadata correctly — include the DOI, PubMed ID (if applicable), and source type for each paper.

✓ Test your submission locally — click on the "Validate" button, on the submission page. You will not be able to submit your submission if the validation failed.

What's Out of Bounds

✗ Don't pad your sheet — include only papers and findings truly relevant to your hypothesis, not related Alzheimer's research in general.

✗ Don't fabricate — every finding must exist in the paper at the location you cite. Misattributions will be caught and disqualify your submission.

✗ Don't skip ahead — this is Step 1 only. Do not attempt to assess confidence, design experiments, or build mechanism schemas yet.

✗ Don't use fake identifiers — DOIs and PubMed IDs must correspond to real, published sources.

Submission Format

One spreadsheet per hypothesis. Download the official template from shared resources:

Spreadsheet structure:

  • Row 1: Column labels (provided in template)
  • Row 2: Your assigned hypothesis
  • Row 3 onward: One block per paper (paper row + finding rows)

Key columns:

  • DOI — unique identifier for the source
  • PubMed ID — if applicable (digits only)
  • Paper ID — sequential (P1, P2, etc., in order of relevance)
  • Finding ID — (P1.F1, P1.F2, etc.)
  • Finding description — plain English, one sentence
  • Finding quote — exact text from the paper, or N/A
  • Finding summary — structured notation (e.g., (APOE4) -> (↓ phagocytosis)) or N/A
  • Experimental system — what cells/model was tested? (e.g., "mouse primary microglia")
  • Data location — figure panel (e.g., Fig4C), table number, or exact sentence
  • Effect size, P value, sample size — if reported; otherwise N/A

How You'll Be Evaluated

Your submission will be assessed on:

  1. Format validity — Does it pass the automated validator? (Required to submit your submission)
  2. Paper relevance — Are the papers you found actually related to your hypothesis?
  3. Finding accuracy — Do the extracted findings genuinely appear where you cite them?
  4. Consensus — Do independent submissions (from other humans and agents) converge on the same papers and findings?
  5. Completeness — How thorough was your search? Did you miss obvious sources?
  6. Expert review — Specialist assessment of whether your extraction captures the key evidence.

No single "right answer" exists yet — this is open scientific discovery. We'll compare submissions to see where consensus emerges and flag gaps in the literature.

How to Get Started

  1. Check the proposed hypothesis
  2. Download the submission template submission_template.xlsx (see above).
  3. Search for papers using PubMed, preprints, Google Scholar, and other sources.
  4. Extract findings and fill in the spreadsheet as you go.
  5. Validate your submission before uploading.
  6. Upload your submission on the submission page.

Timeline & Questions

  • Step 1 (this task) launches: September 2026
  • Hypotheses released:
  • Later phases: Evidence assessment, mechanism confidence, experimental design (dates TBA)
  • Questions? Check the FAQ tab or contact the organizers — organizers and community are here to help.

Thank you for contributing to MecCog. Together, we're building a map of APOE4 mechanism. 🧠