Bio Epidemiological Genomics Amr Surveillance

Version Compatibility

Reference examples tested with: ncbi-amrfinderplus 4.0+, resfinder 4.5+, rgi 6.0+ (CARD 3.3+), abritamr 1.0.14+, staramr 0.10+, hamronization 1.1+, tb-profiler 6.2+, mykrobe 0.13+, mob_suite 3.1+, plasmidfinder 2.1+, MobileElementFinder 1.0+, pandas 2.2+, BioPython 1.84+.

Before using code patterns, verify installed versions match. If versions differ:

• Python: pip show <package> then help(module.function) to check signatures
• CLI: <tool> --version then <tool> --help to confirm flags
• AMRFinderPlus: amrfinder --list_organisms for the current --organism catalogue
• WHO Mtb catalogue: tb-profiler list_db and verify the bundled WHO edition
• AMR database freshness: amrfinder -u (NCBI ReferenceGeneCatalog) and tb-profiler update_tbdb

If a flag or column name does not match (--species vs --organism, barcode_build vs barcode-build), introspect the installed package rather than retrying. AMRFinderPlus, RGI, and ResFinder all renamed columns between major releases.

AMR Surveillance

"What antibiotic-resistance determinants are in this assembly, and what susceptibility do they imply?" -> Combine acquired-gene detection, chromosomal point-mutation calling, mobile-element context, and curated phenotype mapping into a single per-isolate report fit for surveillance or clinical handover. Tool choice is determined by species (Mtb needs TB-Profiler, not AMRFinderPlus), reporting standard (WHO GLASS vs CARD vs in-house), and whether mobility matters (carbapenemase outbreak: yes; routine ESBL screen: less so).

• CLI: amrfinder -n assembly.fa --organism Klebsiella_pneumoniae --plus -- acquired + intrinsic-aware point mutations
• CLI: tb-profiler profile -1 r1.fq.gz -2 r2.fq.gz -p sample -- TB drug-resistance against WHO 2nd-edition catalogue
• CLI: hamronize amrfinderplus --analysis_software_version 4.0.3 --reference_database_version 2025-02-01.1 in.tsv > out.tsv -- normalise output across tools
• Python: pandas to join AMRFinderPlus + RGI + ResFinder + MOB-suite per-isolate for a single decision table

The Single Most Important Modern Insight -- Acquired AMR detection is not species-agnostic

A pan-species AMRFinderPlus run misses chromosomal point-mutation resistance because the point-mutation panels are species-specific and activated only when --organism is set. Running amrfinder -n mtb.fa without --organism returns "no AMR detected" on an XDR-TB genome because AMRFinderPlus has no Mtb organism mode -- the user must switch tools (TB-Profiler or Mykrobe + WHO 2nd-edition catalogue). Similarly, Salmonella gyrA T83I, Klebsiella mgrB inactivation, and E. coli QRDR mutations are silent without --organism Salmonella / Klebsiella_pneumoniae / Escherichia. For any cross-tool surveillance pipeline, AMRFinderPlus with --organism MUST be paired with a species-specific second tool (TB-Profiler for Mtb; ResFinder 4.0 with -s 'species' for PointFinder coverage; hAMRonization to merge). Andersson et al. Nat Rev Microbiol 17:479 (2019) further notes that heteroresistance at 0.1-1% allele frequency is widespread and invisible to default variant callers, compounding the failure mode for any single-tool workflow.

Algorithmic Taxonomy

Decision Tree by Scenario

Methodology evolves rapidly; before a high-stakes outbreak report, web-search "AMRFinderPlus organism modes 2026" and confirm the WHO Mtb catalogue edition currently bundled in TB-Profiler.

AMRFinderPlus With Species Mode

Goal: Produce a per-isolate AMR report including acquired genes, species-specific chromosomal point mutations, stress/virulence elements, and per-hit method and coverage metadata for downstream harmonisation.

Approach: Run amrfinder with -n for nucleotide assembly, --organism to activate the species-specific point-mutation panel, --plus to include stress / virulence / heat / metal, and --report_all when truncated-gene visibility matters. Pin the database via amrfinder -u -> verify --db matches a recorded date.

DB=$(amrfinder -V | grep -i database | awk '{print $NF}')

amrfinder \
    -n assembly.fa \
    --organism Klebsiella_pneumoniae \
    --plus \
    --threads 8 \
    -o sample.amrfinder.tsv

echo "DB version: ${DB}" >> sample.amrfinder.tsv

Column semantics worth inspecting: Element type (AMR / POINT / VIRULENCE / STRESS / etc.), Method (EXACTX / PARTIAL_CONTIG_END / HMM / etc. -- partial-contig hits flag assembly fragmentation), % Coverage of reference sequence, % Identity to reference sequence, Class / Subclass (drug-class summary). Method = PARTIAL_CONTIG_END is the smoking gun for plasmid-context fragmentation -- consider long-read confirmation.

TB-Specific Workflow

Goal: Generate a WHO-catalogue-aligned drug-resistance report for M. tuberculosis covering all 13 first-line + second-line + new/repurposed drugs, with lineage assignment via the Coll/Napier MTBC barcode (90-SNP version) and explicit handling of Group 3 (Uncertain) mutations.

Approach: TB-Profiler primary (it bundles the WHO catalogue and reports the Walker 2022 / 2023 association grouping); Mykrobe as an orthogonal cross-check on any R / XDR call that will drive treatment. NEVER collapse Group 3 mutations to "susceptible" -- the WHO catalogue's tiered scoring is load-bearing for clinical handover.

tb-profiler update_tbdb

tb-profiler profile \
    -1 reads_R1.fq.gz \
    -2 reads_R2.fq.gz \
    -p sample \
    --txt --csv --pdf \
    --dir tbprofiler_out

mykrobe predict \
    --sample sample \
    --species tb \
    --output sample.mykrobe.json \
    --format json \
    reads_R1.fq.gz reads_R2.fq.gz

Allix-Béguec et al. (CRyPTIC) NEJM 379:1403 (2018) established >99% negative predictive value for first-line drugs and is the basis for WGS-only DST policies; the same study showed sensitivity remains lower for bedaquiline, delamanid, linezolid, and clofazimine, so phenotypic DST is still required for second-line and new/repurposed agents in MDR/XDR-TB workups. The WHO 2023 catalogue (Walker et al. Lancet Microbe 3:e265, 2022 for the 2021 edition methodology; 2023 second edition data) grades each mutation Group 1 / Group 2 (Associated -- interim) / Group 3 (Uncertain) / Group 4 (Not Associated -- interim) / Group 5 (Not Associated). Reporting Group 3 as "S" actively misleads clinicians.

hAMRonization for Cross-Tool Reporting

Goal: Convert per-tool AMR output (AMRFinderPlus, ResFinder, RGI, abricate, staramr, ARIBA, TB-Profiler, Mykrobe) to the PHA4GE schema so a multi-laboratory or multi-tool surveillance dataset is comparable.

Approach: Invoke hamronize <tool> per input with mandatory provenance metadata (--analysis_software_version, --reference_database_version, --input_file_name), then hamronize summarize to merge.

hamronize amrfinderplus \
    --analysis_software_version 4.0.3 \
    --reference_database_version 2025-02-01.1 \
    --input_file_name sample.amrfinder.tsv \
    sample.amrfinder.tsv > sample.hamr.tsv

hamronize resfinder \
    --analysis_software_version 4.5.0 \
    --reference_database_version 2024-12-15 \
    --input_file_name sample.resfinder.json \
    sample.resfinder.json > sample.resfinder.hamr.tsv

hamronize summarize -t tsv -o cohort.hamr.tsv per_sample_hamr/*.tsv

Default reporting ontology for public-health output is NCBI ReferenceGeneCatalog. Two consequences: (1) CARD ARO terms need translation; (2) any custom gene additions need a corresponding NCBI accession before they appear in the harmonised stream.

Mobile-Genetic-Element Context

Goal: Determine whether a clinically actionable AMR gene (carbapenemase, mcr, ESBL) sits on a chromosome, a plasmid (and which incompatibility group / cluster), an integrative-conjugative element, or a transposon -- because "gene present" tells the infection-control team nothing about transmissibility.

Approach: Reconstruct plasmids from the assembly with MOB-suite mob_recon; type each plasmid with mob_typer; cross-reference AMR-gene coordinates against the plasmid contig list; annotate IS elements / integrons with MobileElementFinder. For surveillance-grade plasmid resolution, prefer long-read (R10.4.1 Q20+) or hybrid assemblies -- short-read draft assemblies routinely fragment plasmid contigs.

mob_recon \
    --infile assembly.fa \
    --outdir mob_out/ \
    --num_threads 4

mob_typer \
    --infile mob_out/plasmid_AA001.fasta \
    --out_file mob_out/plasmid_AA001.typed.tsv

mefinder find \
    --contig assembly.fa \
    --out mef_out/sample \
    --threads 4

Document MOB-suite version explicitly: v2 and v3 cluster codes are non-interoperable, and v3.1+ adds MGE reporting. Across longitudinal surveillance crossing the v2 -> v3 boundary, the same plasmid receives different cluster IDs and appears spuriously "novel".

Per-Method Failure Modes

AMRFinderPlus run pan-species on a Mycobacterium tuberculosis assembly

Trigger: amrfinder -n mtb.fa without --organism; AMRFinderPlus has no Mtb organism mode in any v4.x release.

Mechanism: The species-specific point-mutation panels are activated only when a recognised --organism is passed. Mtb resistance is overwhelmingly chromosomal point mutation (rpoB / katG / inhA / pncA / embB / gyrA / rrs). With no panel active, none of these are called.

Symptom: Empty AMR table on a phenotypically MDR/XDR isolate; lineage barcode not reported (AMRFinderPlus does not call MTBC lineage).

Fix: Switch tools. Use TB-Profiler (preferred -- bundles WHO catalogue + Coll/Napier barcode) or Mykrobe; the AMRFinderPlus catalogue does not cover MTBC.

OXA-48-like family collapsed to a single bucket

Trigger: Default per-isolate summarisers report the gene family (bla_OXA-48-like) rather than the allele (bla_OXA-244).

Mechanism: OXA-48 (potent carbapenemase), OXA-181 (Thr213Ala -- similar activity, different plasmid), OXA-232 (Arg214Ser -- reduced carbapenemase activity), and OXA-244 (Arg214Gly -- weakest, often phenotypically susceptible to ertapenem) share >95% identity. Pipelines that report family-level summary erase the clinically actionable variant identity.

Symptom: Surveillance report says "OXA-48-like detected", phenotype is borderline ertapenem-susceptible / meropenem-susceptible, clinical team is confused about whether ceftazidime-avibactam is needed.

Fix: Report the allele explicitly; never collapse to family. AMRFinderPlus reports the allele in Gene symbol -- preserve it through hAMRonization rather than summarising.

IS-element-mediated derepression invisible to gene-presence pipelines

Trigger: Phenotypic high-level AmpC hyperproduction in Enterobacter cloacae complex / Citrobacter freundii; or KPC over-expression on Tn4401b (100-bp promoter deletion).

Mechanism: Resistance is mediated by ISEcp1 / IS26 / IS10 insertion upstream of the ampC promoter producing a strong hybrid promoter, or by Tn4401 variant. Gene presence is unchanged. Short-read assemblies fragment IS elements and collapse repeats; promoter context is lost. AMRFinderPlus / RGI / ResFinder report the gene; the regulatory configuration is not annotated.

Symptom: AMR pipeline output identical for wild-type ampC carrier and the hyperproducer; clinical phenotype is divergent.

Fix: Long-read (Nanopore R10.4.1 Q20+ or PacBio HiFi) or hybrid assembly with explicit promoter-context inspection. For Tn4401 variant typing, manual blast of the transposon region. Tools that automate this are emerging but not yet a default in surveillance pipelines.

Heteroresistance below default variant-caller thresholds

Trigger: Routine Illumina WGS-AMR pipeline on a clinical isolate; resistance variant appears at 1-5% allele frequency.

Mechanism: Heteroresistance -- clonal subpopulations carrying resistance at 0.1-1% (Andersson, Nicoloff, Hjort Nat Rev Microbiol 17:479, 2019) -- is widespread and clinically meaningful (treatment failure under antibiotic pressure). Default variant callers (bcftools, lofreq default -q 20, GATK HaplotypeCaller) require minor-allele frequency of typically 10-20% to call. Assembly-based pipelines (AMRFinderPlus on assembly) collapse minor variants entirely.

Symptom: Surveillance pipeline calls "S" but clinical failure under therapy; subsequent sampling reveals high-level resistance.

Fix: For clinically critical drugs, supplement assembly-based AMR with deep-read variant calling at lower MAF thresholds (lofreq with -q 13 -a 0.01), or use targeted deep-amplicon sequencing. TB-Profiler reports allele frequency for resistance calls -- inspect for heteroresistance routinely.