Metal-AMR Site Co-Selection Analysis

Why It Matters

The Atlas currently treats metal-AMR co-selection as plausible but unresolved. This opportunity turns the tension into a concrete analysis: do metal-associated taxa, genes, or environments also show enriched AMR mechanisms after controlling for taxonomy and habitat?

Review Brief

What changed: the analysis now needs to account for prophage density, metal type diversity, soil metal co-contamination, and mobile-element burden in addition to the original metal and AMR profiles.

Why review matters: this opportunity is ready enough to execute, but only if the design distinguishes co-selection from correlated habitat, taxonomy, sampling, and mobile-element effects.

Evidence to inspect:

• Metal-AMR co-selection readiness for unresolved controls.
• AMR Fitness Profiles and Metal Tolerance Scores for reusable inputs.
• prophage_amr_comobilization and microbeatlas_metal_ecology for new covariates.
• ENIGMA/NMDC metadata and soil metal projects for site context.

Questions for reviewers:

• What is the right first benchmark: ENIGMA sites, global MAGs, soil metal datasets, or a smaller matched subset?
• Which null model would convince reviewers that co-selection is not just taxonomy or habitat?
• Should mobile-element burden be modeled before or after metal exposure?
• What negative result would be useful enough to record as a claim-narrowing outcome?

Evidence Base

The analysis can reuse AMR Fitness Profiles, Metal Tolerance Scores, ENIGMA site ecology, and environmental metadata. The key is not to show that metals and AMR both occur in the same broad environment, but to test whether the co-occurrence remains after stronger controls.

Work Package

Build matched site or sample groups by environment, taxonomy, and metadata completeness. Compute AMR burden, metal tolerance score, and mechanism composition for each group. Test whether metal signal explains AMR variation beyond confounders and record where metadata gaps prevent interpretation.

Decision Use

If supported, metal-AMR co-selection becomes a stronger DOE-relevant direction. If not supported, the Atlas should preserve the negative result and narrow co-selection claims to specific taxa, metals, or sites.