# Manuscript readiness and permitted claims

## Computational work completed

Retained-file and sequence audits; identifier validation; descriptive count summaries; full-sequence re-alignment of reported AMP pairs; Phobius predictions; independent precursor-model prediction across all predicted proteins; mature-model eligibility and nine-record Macrel check; fresh DRAMP database search; Swiss-Prot functional/homology searches of proteins and transcripts; lineage-based first-pass taxonomic review; UniVec vector/adaptor review; Hymenoptera-specific protein BUSCO; deduplicated candidate review; sample metadata and methods/results draft; figures and reproducibility/checksum package.

## What the results support

- Reporting the retained assembly, predicted proteins and delivered per-sample counts with explicit provenance and limitations.
- Describing homology-based functions and mixed taxonomic signals as computational evidence.
- Reporting exploratory AMP scores and conflicting evidence without asserting activity, novelty or confirmed ant origin.
- Within-colony descriptive observations from three worker ants. There is no treatment contrast or independent colony replication.

## What they do not support

- A confirmed novel ant-derived AMP, antimicrobial activity, safety, secretion, mechanism or therapeutic claim.
- Treating whole-body microbial homologues as proven laboratory contamination, or treating no-hit sequences as clean host sequence.
- A complete host-only assembly, validated taxonomic partition, complete expression matrix, TPM values or differential-expression conclusions.
- A causal or population-level claim from this single-colony design.

## Evidence requiring resolution before final manuscript submission

1. **Specimen identity/context:** the author corrected the species to *Diacamma assamense* on 4 November 2026 and reports identification by a taxonomist. The earlier locality/range concern related to the superseded label. Original vendor names are retained for source provenance; the corrected name is used in the report and manuscript. Taxonomist name and voucher details are omitted at the author's request.
2. **Library and mapping methods:** extraction, library preparation/strandedness, mapping reference, exact commands and count-selection rules are not established. Obtain authoritative methods or explicitly limit the paper to a reanalysis of retained deliverables.
3. **Assembly interpretation:** substantial microbial homology, vector-score flags and limited protein BUSCO completeness require manual source review before host-specific biological conclusions or automatic filtering. A large comprehensive nucleotide-taxonomy screen (e.g. FCS-GX) has not been run; do not describe the current review as a contamination-free certificate.
4. **Data availability:** the original raw reads were unavailable at the supplied path when checked; restore them and arrange appropriate archival deposition/accessions if pursuing a transcriptome resource publication. No accession has been fabricated or submitted.
5. **Experimental validation:** no laboratory access is available. Experimental validation cannot be completed computationally. A computational paper may transparently report predictions; experimentally demonstrated antimicrobial claims require actual measurements. A prospective validation design is provided separately.

## Current candidate interpretation

The original 11-record secretion-based tier is historical and is superseded by the integrated review. No original vendor candidate reaches the exploratory ampir score cutoff of 0.8; this is threshold-dependent, not proof that none is an AMP. The three new high-score/complete-ORF/SP/no-TM records encode one peptide sequence with a best phage-spanin reference match. The sole record overlapping the high-score and fresh DRAMP-match sets has a partial ORF and best Swiss-Prot homology to human S100-A9. Neither observation establishes ant-derived AMP identity.

The deliverables are ready for reviewing and drafting a computational assessment. Submission-ready biological claims still depend on the external evidence above. No experiments, restored reads, deposition accessions, species re-identification or completed host partition are claimed.