Skip to content

output-structure.md

Latest commit

 

History

History
76 lines (58 loc) · 4.29 KB

output-structure.md

File metadata and controls

76 lines (58 loc) · 4.29 KB
Raw

comMS documentation: Commands · Configuration · Configuration reference · Output structure · README

Output structure

This page describes the directory structure that comMS uses to write results and logs. For how the output root is chosen, see experiment context.

Contents

Results layout

All outputs are written under the experiment root, inside comms/results/. Each command has its own subdirectory:

<experiment_root>/
  └─ comms/
     └─ results/
        ├─ convert/                # indexed .mzML files
        ├─ index/                  # Crux tide-index output
        ├─ search/                 # Crux tide-search target PSM files
        ├─ rescore/                # Percolator rescored PSM files
        │  └─ <organism>/          # per-organism split PSMs and assign-confidence output (only in multi-species analysis mode)
        ├─ lfq/                    # MS1 label-free quantification output
        │  └─ <fraction>/          # FlashLFQ output for each fraction
        ├─ quantify/               # dNSAF spectral-counts output
        └─ report/                 # report output
           ├─ index.md             # index of produced files, parameters, and section status
           ├─ qc/                  # quality-control plots and spreadsheet
           ├─ pca/                 # PCA and dendrogram plots
           ├─ da/                  # differential abundance plots and spreadsheet
           ├─ secondary_species/   # secondary-species plots and spreadsheet
           ├─ concordance/         # LFQ vs dNSAF concordance (only if --lfq-dir is provided)
           └─ ev_markers/          # MISEV2023 marker-category heatmaps (only if run via --section/--all)

If a command's output directory already exists, comMS does not overwrite it. Instead it adds an incremental suffix, for example search-1/, then search-2/, so earlier results are preserved.

comMS commands which rely on the output of other commands will resolve the input directory by following the above convention, unless overridden on the command-line. It will not, however, use the latest version of an output directory (i.e. search-1/) but instead default to the expected <command>/.

Logging

comMS logs to both the terminal and a file.

Log levels

Errors, warnings, and high-level information are always shown. Verbosity flags enable more detail:

Level Enabled with Used for
debug -vv Detailed program state for debugging: the command invoked, paths scanned, resolved parameters, and per-item detail
progress -v Step-wise progress: per-file and per-stage markers
info always on High-level overview: processing counts and final results
input always on (interactive sessions only) Confirmation prompts and their answers, for example when config or experiment need to create a file or confirm a destructive action
warn always on A part of a command did not succeed but comMS continued, for example a single item failed, an optional input was missing, or a fallback value was used
error always on A part of a command did not succeed and comMS could not continue

-vv implies -v. The input level only prompts interactively when running in a terminal (stdin is a TTY); otherwise it falls back to the flag's default silently.

Log files

comMS logs to standard output (the terminal you ran the command from) and to a file named after the command (comms convert, for example, writes to convert.log). The log file is saved alongside that command's other output, so running the same command again does not overwrite an earlier log.

comMS documentation: Commands · Configuration · Configuration reference · Output structure · README