Transparent target prioritization in Leishmania major and a fully auditable AutoDock Vina workflow for the platelet-activating factor acetylhydrolase LmjF.35.3020 (UniProt E9AFC6)

Leishmaniasis remains difficult to treat because the available medicines carry toxicity, administration, cost and susceptibility limitations. We report an auditable framework for prioritizing candidate drug targets in Leishmania major with a fully documented docking case study for LmjF.35.3020, a platelet-activating factor acetylhydrolase (PAF-AH; UniProt E9AFC6). A master table of 6,457 parasite proteins was audited for annotation coverage and multi-criterion scores. Coverage was heterogeneous (InterPro 54.11%, Gene Ontology 29.72%, KEGG 2.85%), a composite score was available for 6,266 proteins (97.04%), and the 40-protein shortlist was strictly nested within the 52-protein candidate export. The composite score was recovered and regenerated exactly for all 6,266 scored proteins as an unweighted sum of five components; two components are misnamed, so InterPro annotation is counted twice and predicted essentiality is an upper-quartile ortholog-count cut rather than genetic evidence. Rank order is robust to reweighting (Spearman rho >= 0.979) but the leading ten proteins are not, and the exported tiers are curated rather than score-ranked. Re-annotation placed LmjF.35.3020 in the PAF-AH family (InterPro IPR016715) on an alpha/beta-hydrolase fold (IPR029058), and receptor geometry confirmed a complete Ser266-His390-Asp299 catalytic triad within a single GXSXG nucleophile elbow, enclosed together with all 28 predicted pocket residues by the 30 A docking grid. BLASTP against a human reference proteome returned PAFAH2 (34.7% identity over 42.3% of the query) and PLA2G7 (28.3% over 68.0%), so the target passes the predeclared host-homology rule, as do all 40 shortlisted proteins; the catalytic serine is nonetheless conserved in PLA2G7, which a global filter cannot detect. A 77-ligand AutoDock Vina v1.2.5 screen, drawn from a documented 1,514-compound cascade, gave predicted best-pose scores from -10.390 to -4.604 kcal/mol (median -7.18 kcal/mol). Without decoys or counter-docking these rank poses and are not evidence of binding.