andes/config.example.yaml at main · bigbio/andes · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
# andes run configuration (YAML) — pass with `andes --config this-file.yaml`.
#
# Every key is OPTIONAL; omitted keys keep their built-in default. An explicit
# CLI flag ALWAYS overrides the value here (precedence: CLI flag > config > default),
# so this file sets the baseline and you tweak individual runs on the command line.
#
# Sections are grouped by experimental design (see DOCS §1a / §1b). Tolerances,
# ranges, and enums are written exactly as on the CLI (e.g. "20ppm", "2..5", auto).
# Unknown keys are a hard error, so typos never silently no-op.

# ── Inputs / outputs ────────────────────────────────────────────────────────
io:
  spectrum: [sample.mzML]        # one or more files (mzML / MGF / .raw / .d)
  database: human.fasta          # target-only FASTA (decoys generated per `decoys:`)
  output_pin: out.pin            # Percolator PIN (required unless rescoring in-process)
  # output_tsv: out.tsv          # optional human-readable PSM table
  # output_parquet: out.idparquet

# ── Core search ─────────────────────────────────────────────────────────────
search:
  precursor_tol: 20ppm           # e.g. 20ppm or 0.02da
  charge: "2..5"                 # fallback charge range when the spectrum omits it
  isotope_error: "-1..2"
  enzyme: trypsin                # trypsin | lysc | gluc | ... | nonspecific | gluc,trypsin
  enzyme_specificity: fully      # fully | semi | non-specific
  max_missed_cleavages: 1
  min_length: 6
  max_length: 50
  mods: mods.txt                 # Java-format mods.txt (TMT/iTRAQ/phospho/etc.)
  # max_mods: 3
  # min_peaks: 10                # advanced
  # top_n: 10                    # advanced

# ── Scoring / model / runtime ───────────────────────────────────────────────
scoring:
  score: auto                    # auto | strong | rank
  precursor_cal: auto            # off | auto | on
  threads: 8
  # fragmentation: auto          # MGF only (no metadata): auto | CID | ETD | HCD | UVPD
  # protocol: auto               # auto | TMT | iTRAQ | phospho | ...
  # fragment_tol_ppm: 20         # MGF only, mutually exclusive with fragment_tol_da
  # model_store: models.parquet  # custom / trained model store
  # model: hcd_qexactive_tryp    # force a specific model id

# ── Decoys & FDR strategy ───────────────────────────────────────────────────
decoys:
  strategy: reverse              # reverse | shuffle | none (input already has decoys)
  prefix: XXX_
  # suffix: _rev                 # for OpenMS/quantms-style suffixed decoy DBs

# ── Experiment modes (each opt-in; sub-knobs are advanced) ───────────────────
chimeric:
  enabled: false                 # co-isolated / co-fragmented peptide cascade
  # max_coisolated: 4
  # max_kl: 0.3

refine:
  enabled: false                 # PTM-discovery second pass
  # config: refine_tiers.yaml
  # select_psm_fdr: 0.01

rescoring:
  enabled: false                 # run Percolator in-process (else feed out.pin to Percolator)
  # fdr: 0.01                    # q-value threshold for filtered output
  # native: false                # non-production built-in GBDT rescorer

glyco:
  enabled: false                 # intact N-glycopeptide search
  # --- advanced tuning (validated defaults; see DOCS §9) ---
  # backbone_top_k: 150
  # gp_k: 10.0                   # glycan-Y ladder weight
  # gp_cz: 15.0                  # ETD c/z hyperscore weight (inert on HCD/CID)
  # hcd_pair: true               # DEFAULT ON: generate from the paired HCD scan, score c/z on ETD
  #
  # The remaining glyco flags (glycan list, taxon, NeuGc, emission floors, ETD c/z
  # knobs, retrieval tolerance, ...) are CLI-only; see DOCS.md §9 for the full table.
  # gp_j: 5.0
  # gp_h: 1.0
  # pf_charge: 2
  # max_pf: 1024