Output files and columns

GVClass writes a run-level summary plus per-query artifacts to the output directory (<query_dir>_results by default, or the path passed to -o). The flags that add or remove files are listed in the CLI reference.

Files

Run-level files

Written to the root of the output directory.

File	Contents
gvclass_summary.tsv	Main summary table, one row per query, tab-separated.
gvclass_summary.csv	Same content as gvclass_summary.tsv, comma-separated.
gvclass_summary.extended.tsv	Always-on per-contig contamination diagnostics, one row per query, tab-separated.
gvclass_summary.extended.csv	Same content as the extended TSV, comma-separated.

Per-query files

Written for each input query.

File	Contents
<query>.final_summary.tsv	Single-query row in the 44-column main schema, used to rebuild the run summary on --resume.
<query>.summary.tab	Legacy per-query summary table.
<query>.tar.gz	Bundled per-query artifacts.
<query>.SUCCESS	Resume sentinel; its presence marks the query complete for --resume.
<query>.contamination_candidates.tsv	Suspicious contigs; written only when estimated_contamination is at least 10, contamination_type is interpretable, and suspicious contigs are found.

Species-tree files

Written only with --species-tree. See Build a species tree and the species tree explanation.

Path	Contents
species_tree/<query>/<query>.treefile	Concatenated-marker species tree for the query.
species_tree/<query>/<query>.partitions.txt	Per-marker partition definitions for the query supermatrix.
species_tree/<query>/species_tree_taxonomy.tsv	Taxonomy and full-precision nearest-reference distances for the query species tree.
species_tree/combined.*	Combined tree across all queries; written only with --species-tree-combined.
species_tree/_combined/<panel>/	Per-panel combined trees for multi-domain batches; written only with --species-tree-combined.

Note

Without --species-tree, no species_tree/ directory is produced and the four species_tree_* summary columns are nd.

Summary columns

gvclass_summary.tsv and gvclass_summary.csv share the same 44 columns in this order.

Column	Description
query	Input filename for the query.
taxonomy_majority	Full lineage from the per-marker single-gene tree nearest-neighbor majority vote.
species_tree_nn_taxonomy	Lineage of the nearest reference in the concatenated-marker species tree; nd unless --species-tree was set.
taxonomy_confidence	high when every emitted rank cleared its distinct-marker threshold, otherwise one or more of low_support, reduced_fastmode, no_support.
capsid_group	Unified capsid-type tally as label:count across the Nucleocytoviricota, Mirusviricota, and Bellas & Sommaruga capsid groups.
species	Species-rank call with per-taxon counts.
genus	Genus-rank call with per-taxon counts.
family	Family-rank call with per-taxon counts.
order	Order-rank call with per-taxon counts.
class	Class-rank call with per-taxon counts.
phylum	Phylum-rank call with per-taxon counts.
domain	Domain-rank call with per-taxon counts.
avgdist	Average tree distance to the reference neighbors.
order_dup	Average copy number of expected order-level markers; elevated values indicate duplicated, chimeric, or mixed bins.
estimated_completeness	Estimated percent of the expected genome recovered for the assigned lineage; the only completeness field in the main table.
completeness_model_reliability	advisory_only, moderate, or high, from the per-order model hold-out R^2; a property of the model, not the genome.
estimated_contamination	Output of the trained extra_trees_v1 contamination model and the primary contamination estimate.
contamination_type	Likely contamination source (clean, cellular, mixed_viral, phage, duplication, uncertain), populated when estimated_contamination is at least 10.
gvog4_completeness	Distinct core NCLDV GVOG4 markers present, as n/4.
gvog4_dup	GVOG4 duplication factor (total marker hits / distinct markers present).
gvog8_completeness	Distinct core NCLDV GVOG8 markers present, as n/8.
gvog8_dup	GVOG8 duplication factor.
busco_completeness	Eukaryotic BUSCO single-copy markers present, as n/255.
busco_dup	BUSCO duplication factor; elevated values indicate cellular (eukaryote) carry-over.
cog_completeness	Universal COG (UNI56) single-copy markers present, as n/56.
cog_dup	COG duplication factor; elevated values indicate cellular (prokaryote) carry-over.
mrya_completeness	Mryavirus markers present, as n/6.
mrya_dup	Mryavirus duplication factor.
phage_completeness	Phage (geNomad) markers present, as n/20.
phage_dup	Phage duplication factor.
ncldv_mcp_total	Count of NCLDV-specific major capsid protein (MCP) markers.
vp_completeness	Virophage core markers present (MCP, Penton, ATPase, Protease), as n/4.
vp_mcp	Count of virophage MCP hits.
plv	Count of A32 (PLV_PC_054) proteins placing with PPV references; 0 for ordinary NCLDV.
mirus_completeness	Mirusviricota core markers present (MCP, ATPase, Portal, Triplex), as n/4.
contigs	Number of contigs in the query.
LENbp	Total length in base pairs.
GCperc	GC content as a percentage.
genecount	Number of predicted genes.
CODINGperc	Coding density as a percentage.
ttable	Genetic code used for gene calling; no_fna for protein inputs.
species_tree_nn_genome	Nearest-reference genome id from the species tree; nd without --species-tree.
species_tree_nn_distance	Distance to the nearest reference in the species tree; nd without --species-tree.
species_tree_clade_id	Clade id of the nearest reference in the species tree; nd without --species-tree.

The per-contig diagnostic columns (cellular_coherent_*, cellular_lineage_purity_median, viral_bearing_contig_count, contig_attribution_mode) appear only in the extended table.

For how to read completeness, contamination, and duplication, see Quality metrics. For the marker panels behind the completeness columns, see Markers.