Files
- Island synteny viewer
- Assembly quality QC
- frequency_table (tsv.gz)
- clade_assignments (tsv.gz)
- strain_inventory (tsv.gz)
- significant_islands (tsv.gz)
- island_pfam_enrichment (tsv.gz)
- pangenome_openness (tsv.gz)
- per_strain_summary (tsv.gz)
- classification_counts (tsv.gz)
- island_size_distribution (tsv.gz)
- marker_summary (tsv.gz)
- islands_with_domains (tsv.gz)
- assembly_quality_vs_content (tsv.gz)
- assembly_quality_correlations (tsv.gz)
- diagnostics (tsv.gz)
- group_class_overlap (tsv.gz)
Pipeline diagnostics
- rescue_redundancy [TRIGGERED]: 15128104/19521312 (77.5%) rescuable cells overlap a predicted gene already assigned to a different family in that strain -- the same locus double-counted, not a genuine annotation dropout (issue #133).
- prune option: do not enable rescue (--pangenome_rescue_enable false); see clustering sweep #132 before re-enabling
- prune option: verify --pangenome_rescue_structural_filter is enabled (default true) -- it is the mechanism that removes these cells, not merely detects them
- assembly_quality_confound [TRIGGERED]: max |rho| = 0.44 (rho = -0.4404, accessory_present vs n50 (partial, controlling total_length), n = 169 strains); threshold 0.3. See assembly_quality_report.md (issue #130). Between-strain accessory-content comparisons are confounded by assembly fragmentation on this dataset.
- prune option: exclude or flag the most fragmented assemblies (highest n_contigs in assembly_quality_vs_content.tsv) and rerun
- prune option: report accessory-content comparisons with assembly quality as a covariate, not as raw per-strain counts
- clade_structure_validity [not computed]: trips when: PCoA1 variance share < ~50%, split-half ARI < 0.8, or between/within clade distance ratio < ~2. Not yet computed by this pipeline -- see #131.
- gain_loss_skew [not computed]: trips when: gain:loss ratio > 20:1 or ambiguous-polarity fraction < 2%. Not yet computed by this pipeline -- see issue #134's calibration reference.
Pangenome report
Definitions
- Gene family: a cluster of proteins from all strains at the tier-1 clustering identity and coverage. A family is present in a strain when a member protein is annotated there, or (with rescue on) when a TBLASTN hit marks it
genome_only. - Frequency class (computed separately for each group, over that group's representative strains):
singleton= one representative strain; otherwise by the fraction of representatives:core>= 0.95,soft_core>= 0.90,shell>= 0.15,cloudbelow 0.15 (defaults; the run parameters set them). A family in no representative getsnonrep_only(only in near-identical strains that were dereplicated away),outgroup_onlyoringroup_only(only in the other group), orabsent. - Representative strain: one strain kept per group of near-identical strains (Mash dereplication). Counts of "strains" that include non-representatives are said so.
- Accessory island: a run of consecutive non-core genes in one strain that contains at least one significantly co-occurring pair that is also physically linked. The islands are not tested as units.
Pangenome composition
Ingroup (26047 gene families)
- core: 6846 (26.3%)
- soft_core: 610 (2.3%)
- shell: 4025 (15.5%)
- cloud: 6764 (26.0%)
- singleton: 7802 (30.0%)
Not counted in this group's classes:
- nonrep_only: 1406
- outgroup_only: 20327
Outgroup (32609 gene families)
- core: 6357 (19.5%)
- soft_core: 636 (2.0%)
- shell: 4628 (14.2%)
- cloud: 11894 (36.5%)
- singleton: 9094 (27.9%)
Not counted in this group's classes:
- nonrep_only: 3591
- ingroup_only: 11580
Pangenome openness
Heaps' law fit: κ=9019.9, γ=0.263 (R²=0.997) -- pangenome is open (γ > 0; γ = 1 would mean every new strain adds the same number of new families). The curve counts every strain in the presence matrix, including near-identical strains and the outgroup.
Extrapolated asymptotic core-genome size: 295 families.
Per-strain summary
Families per strain: min=8972, median=9586, max=10324 (n=529 strains). See per_strain_summary.tsv for outliers.
Outlier strains (singleton-count modified z-score beyond threshold): B3476, h5384, 485B-1_L_OLD_CPA0023, 4545-MICE, CA14, GT-101, UTAH_23742X6, FMC778, GT-129, GT-139, VFC084_CPA0067, GT-160, SOIL_604-1_L_OLD_CPA066, GT-133, B11863, Michoacan_2, VFC083, GT-100, CA1, J_TORRES, B11002, Cocci_1400035797, UTAH_20380X2, COCPO_34698, COCSP-3505, COCPO_195881, SA14, NACVFR_4542, B14286, B11518, GT-147, COCPO_4545, NACVFR_2566, B1249_Guatemala, 730333_Guatemala, B0858-Guatemala, 730332_Guatemala, 409B-0_L_OLD_CPA0011, UTAH_9443
Ingroup vs outgroup content
Each gene family's class in the ingroup against its class in the outgroup (counts; colour is log-scaled).
Families in both groups, split by ingroup class, and families found in one group only.
Every non-empty group-class combination, largest first.
Counts: the run's group_class_overlap.tsv table.
Accessory islands
11968 accessory islands found (built from adjacency of non-core genes, gated by containing at least one FDR-significant physically-linked pair). The islands themselves are not tested as units.
Top islands (by size):
7733 single-strain islands excluded (present in one strain only -- strain-private content, which dominates the size ranking).
Pair classification breakdown
- trans_unconfirmed: 68463
- unexplained_physical: 5048
- ambiguous_linkage: 2009
- insufficient_data: 1539
Trans-module genomic clustering
No trans modules to score.
Pfam domain enrichment
| Domain | Fisher p | FDR q |
|---|---|---|
| DUF3435 | 5.09e-18 | 4.73e-15 |
| DUF3723 | 3.26e-12 | 1.51e-09 |
| APH | 1.86e-09 | 5.75e-07 |
| Pkinase_fungal | 7.88e-08 | 1.83e-05 |
| Pkinase | 1.58e-06 | 2.84e-04 |
| DUF3723_N | 1.84e-06 | 2.84e-04 |
| DUF7924 | 2.42e-06 | 3.20e-04 |
| NAD_binding_6 | 1.66e-05 | 1.93e-03 |
| DUF7779 | 7.22e-05 | 7.44e-03 |
| ABC_tran | 2.94e-04 | 2.73e-02 |
| Myb_DNA-bind_6 | 5.60e-04 | 4.69e-02 |
| p450 | 6.06e-04 | 4.69e-02 |