The Evidence Layer for Scientific AI

 Quick Analysis Plan Copied

 Long Analysis Plan

Evidence Base for the Convergent-Node Hypothesis

What the supplied records support: MprF is the best-quantified node. In S. aureus Newman ΔmprF complemented strains, single hot-spot SNPs (T345A, S295L, L826F) raised daptomycin MICs from 0.5 mg/L to 1–2 mg/L while increasing hNP-1/LL-37 survival modestly; dual SNPs (S295L+L826F, T345A+L826F) dropped MICs to 0.38 mg/L, collapsed CAMP survival to 6.8–19.7%, and increased daptomycin binding (0.84–0.98 vs 0.59–0.72 µM) — the authors interpret this as an L-PG-dependent charge-repulsion ceiling plus fitness cost Dual mprF SNPs collapse daptomycin resistance. Independent FRET/planar-bilayer work in S. aureus SA113 confirms MprF-mediated resistance coincides with altered peptide interaction with L-PG-containing membranes, though with peptide-specific behaviors MprF L-PG membrane peptide interaction. For lipid A acylation, K. pneumoniae B5055ΔlpxM (under-acylated lipid A) showed 16-fold lower polymyxin B MIC (0.125 vs 2 mg/L), ~4-fold lower polymyxin E MIC (0.25 vs 4 mg/L), ~2–5-fold lower binding Kd, and ~2.5-fold lower outer-membrane permeabilization EC50 Under-acylated lipid A polymyxin susceptibility.

What is missing from the supplied records: dlt D-alanylation, efflux pumps (e.g., CrrAB/TagF-linked pathways), and physical sequestration have no claim-level evidence here — BGPT inference only from the query premise; adjuvant-target ranking across these nodes cannot be evidence-graded. The DHHC acyltransferase structural record (PDB 6BMN catalytic triad Asp153-His154-Cys156; four-TM hydrophobic cavity as a fatty-acyl "molecular ruler") offers a transferable structural framework for inhibitor design against membrane acyltransferases, though it is eukaryotic, not bacterial lipid A machinery DHHC acyltransferase structural mechanism.

Recommended Bioinformatics/Analysis Plan

• Homology + structure: Build AlphaFold/3Dmol models of bacterial MprF and lipid A transferases (ArnT/PagL/lpxM family); anchor on 6BMN catalytic geometry for pocket enumeration.
• Quantitative resistance modeling: Fit MIC vs genotype (single vs dual mprF SNPs) with epistasis terms; test whether the L-PG-charge model predicts the observed dual-SNP collapse.
• Binding thermodynamics: Recompute ΔΔG of polymyxin interaction with tetra- vs under-acylated lipid A from the reported Kd values; cross-validate with docking (DiffDock) against membrane-representative bilayers.
• Adjuvant target scoring: Rank MprF, lpxM/ArnT, and dlt enzymes by (i) genetic tractability of resistance (the dual-SNP cost data suggest a low evolutionary ceiling for MprF mutational escape), (ii) structural druggability, and (iii) cross-species conservation — efflux-pump ranking waits on additional evidence.

Known Unknowns and Blindspots

All mprF data derive from one MSSA background (Newman) with plasmid complementation — chromosomal SNPs and MRSA backgrounds are untested Study limitations single background. The K. pneumoniae lipid A work used only two strains Two-strain limitation. Falsification: demonstrating that MIC changes track with non-L-PG/lipid A mechanisms (e.g., cell-wall thickening alone) in chromosomal mutants would weaken the convergent-surface-charge model.