[
 {
  "label": "Organism-label class",
  "field": "sp_nonmeta_class",
  "level": "run",
  "description": "Class of the run's ENA organism/scientific_name label: not_flagged (metagenome/community term), named_human_host ('Homo sapiens' — routine submitter label, NOT a quality signal), named_nonhuman_host (a non-human animal species, e.g. Mus musculus), named_microbe_or_other (a single named microbe or other non-community organism). Sample-level column = highest-priority class over the sample's runs (nonhuman_host > microbe_or_other > human_host > not_flagged).",
  "used": true
 },
 {
  "label": "Non-metagenome organism (strict)",
  "field": "sp_flag_non_metagenome_strict",
  "level": "run",
  "description": "Sandpiper rule: organism name names a single specific species (including 'Homo sapiens'); names containing 'metagenome' or community terms (uncultured, environmental sample, enrichment/mixed culture, microbial community, consortium, microbiome) and placeholders (bacterium, unidentified, archaeon, prokaryote, eukaryote, organism, microorganism) are not flagged. Kept for reproducibility; noisy because of the human-host label.",
  "used": true
 },
 {
  "label": "Non-metagenome organism (loose)",
  "field": "sp_flag_non_metagenome_loose",
  "level": "run",
  "description": "Sandpiper rule: organism name lacks 'metagenome' and any community term; placeholders still count as flagged.",
  "used": true
 },
 {
  "label": "Non-metagenome label (curated)",
  "field": "sp_flag_non_metagenome",
  "level": "sample",
  "description": "True when sp_nonmeta_class in {named_nonhuman_host, named_microbe_or_other} over the sample's runs (v1.2.1; v1.2.0 used the strict rule and was dominated by the 'Homo sapiens' label).",
  "used": true
 },
 {
  "label": "Synthetic",
  "field": "sp_flag_synthetic",
  "level": "run+sample",
  "description": "Organism name contains 'synthetic' or 'simulat', or library_source = SYNTHETIC. Sample level: any run.",
  "used": true
 },
 {
  "label": "RNA / non-DNA (strict)",
  "field": "sp_flag_rna_strict",
  "level": "run",
  "description": "library_strategy in {RNA-Seq, miRNA-Seq, FL-cDNA, ssRNA-seq, ncRNA-Seq, RIP-Seq, Ribo-seq, EST} or library_source in {METATRANSCRIPTOMIC, TRANSCRIPTOMIC, TRANSCRIPTOMIC SINGLE CELL}.",
  "used": true
 },
 {
  "label": "RNA / non-DNA (loose)",
  "field": "sp_flag_rna_loose",
  "level": "run",
  "description": "As strict, plus library_source in {OTHER, SYNTHETIC}.",
  "used": true
 },
 {
  "label": "RNA / non-DNA",
  "field": "sp_flag_rna",
  "level": "sample",
  "description": "Any run fires the strict RNA rule.",
  "used": true
 },
 {
  "label": "Low complexity",
  "field": "sp_flag_low_complexity",
  "level": "run+sample",
  "description": "Sandpiper low-complexity rule on the profile (few taxa dominate). Expected for Bifidobacterium-/Enterobacterales-dominated neonatal stool; display only, never a triage input.",
  "used": true
 },
 {
  "label": "Low depth",
  "field": "sp_low_depth",
  "level": "run+sample",
  "description": "Root coverage < 2 genome equivalents (sample level: summed over profiled runs). Such samples are excluded from study panels and study medians (v1.2.1).",
  "used": true
 },
 {
  "label": "Read-fraction warning",
  "field": "sp_flag_readfraction_warning",
  "level": "run+sample",
  "description": "Sandpiper reports a warning on the prokaryotic read fraction (sp_spf) for the run.",
  "used": true
 },
 {
  "label": "Predicted ecological (host-or-not)",
  "field": "sp_predicted_ecological",
  "level": "run",
  "description": "Sandpiper's sp_prediction / sp_host_or_not says the community looks environmental rather than host-associated.",
  "used": true
 },
 {
  "label": "Low known-species fraction at high depth",
  "field": "sp_low_known_species_high_depth",
  "level": "run",
  "description": "Known-species fraction low despite adequate depth — novel community or off-target sample; study-level fraction is frac_low_ksf_high_depth_profiled.",
  "used": true
 },
 {
  "label": "Runs flagged",
  "field": "sp_frac_runs_flagged",
  "level": "study",
  "description": "Fraction of the study's runs with sp_nonmeta_class in {named_nonhuman_host, named_microbe_or_other} or synthetic or RNA (strict). The 'Homo sapiens' label is not counted.",
  "used": true
 },
 {
  "label": "Low-complexity fraction",
  "field": "sp_frac_low_complexity_profiled",
  "level": "study",
  "description": "Fraction of profiled samples with sp_flag_low_complexity.",
  "used": true
 },
 {
  "label": "Low-depth fraction",
  "field": "sp_frac_low_depth_profiled",
  "level": "study",
  "description": "Fraction of profiled samples with sp_low_depth.",
  "used": true
 },
 {
  "label": "Auditor attention",
  "field": "sp_auditor_attention",
  "level": "study",
  "description": "Hint that a study deserves a human look (reason in sp_auditor_reason); never a verdict.",
  "used": true
 },
 {
  "label": "Domain-only (GTDB)",
  "field": "domain_only_gtdb",
  "level": "run",
  "description": "Every classification in the condensed GTDB profile stops at domain. Not shipped in this package (all catalog profiles reach phylum or deeper for the run set used).",
  "used": false
 },
 {
  "label": "Domain-only (GlobDB)",
  "field": "domain_only_globdb",
  "level": "run",
  "description": "GlobDB-scheme profile flag. UNUSED: the catalog uses GTDB profiles only.",
  "used": false
 },
 {
  "label": "Domain-only (both)",
  "field": "domain_only_both",
  "level": "run",
  "description": "UNUSED: requires the GlobDB profile, which the catalog does not use.",
  "used": false
 }
]