Blind search mode parameter optimization

blind-search-mode-parameter-optimization

Summary

A mode of natural product database search that relaxes structural constraints to identify peptides with arbitrary post-translational modifications (PTMs) without prior knowledge of modification types. This skill is essential when screening for novel RiPP variants that fall outside the scope of curated PTM databases.

When to use

When you have tandem mass spectra from ribosomally synthesized peptides (RiPPs) and suspect the presence of unknown or non-standard post-translational modifications that would be missed by standard database search modes constrained to known modification classes. Apply this when initial targeted searches (e.g., MetaMiner with default lantibiotic class) yield incomplete hit coverage or when exploring spectral regions with unidentified high-confidence peaks.

When NOT to use

• Input spectra are from non-peptidic natural products or metabolites outside the RiPP class — use Dereplicator+ instead for general metabolite identification.
• The modification landscape is well-characterized and constrained to known PTM types (e.g., standard lanthipeptides) — standard (non-blind) mode is more computationally efficient and has better statistical calibration.
• Spectral data quality is poor (low signal-to-noise, insufficient fragmentation) — blind mode will amplify false positives without targeted constraints.

Inputs

• tandem mass spectra in MGF or mzML format (centroided)
• RiPP precursor sequence file (FASTA format)
• RiPP structure database

Outputs

• significant_matches.tsv (scan identifiers, match scores, p-values, false discovery rates)
• cross-comparison table of blind vs. standard mode hits
• differential detection summary (blind-mode-only, standard-mode-only, common identifications)

How to apply

Execute the database search pipeline (e.g., MetaMiner) twice: first with standard parameters to establish a baseline of known PTM hits, then re-execute with the --blind flag enabled to search for arbitrary modifications without constraining to predefined PTM types. Parse and cross-compare the significant_matches.tsv outputs from both runs, recording scan identifiers, match scores, p-values, and false discovery rates. Tabulate tool-specific detections and common identifications between modes to identify modifications and hit patterns unique to blind mode, using this differential analysis to distinguish genuine novel modifications from noise and assess mode-dependent sensitivity.

Related tools

• MetaMiner (Primary pipeline for blind-mode PTM-agnostic RiPP database search; accepts --blind flag to relax structural constraints) — https://github.com/mohimanilab/MetaMiner
• ProteoWizard msconvert (Converts spectra in non-native formats (mzXML, mzData, mzML) to MGF for compatibility with blind search) — http://proteowizard.sourceforge.net/
• NPDtools 2.5.0 (Toolkit containing MetaMiner and related database search pipelines (Dereplicator, VarQuest, Dereplicator+); version 2.5.0 includes blind-mode support) — https://github.com/ablab/npdtools

Examples

python metaminer.py test_data/metaminer/msms/ -s test_data/metaminer/fasta/ -o metaminer_outdir && python metaminer.py test_data/metaminer/msms/ -s test_data/metaminer/fasta/ --blind -o metaminer_blind_outdir

Evaluation signals

• Blind mode outputs additional valid matches (non-zero rows in significant_matches.tsv) not present in standard mode, with p-values and FDR consistent with statistical significance thresholds
• Scan identifiers and match scores from blind mode are reproducible across independent executions
• Cross-comparison reveals distinct modification patterns in blind-mode-only hits (e.g., mass shifts not in the standard PTM dictionary) with fragmentation patterns consistent with real PTMs
• False discovery rate (FDR) in blind mode remains below acceptable threshold (e.g., < 0.05 or < 0.10 depending on downstream validation), assessed via spectral quality and chemical plausibility of identified modifications
• Blind mode hits can be validated by independent spectral networking or metabologenomic integration (e.g., presence of corresponding biosynthetic gene clusters) where available

Limitations

• Blind mode substantially increases computational cost and runtime compared to standard (constrained) mode, as it searches over a vastly larger space of possible modifications.
• False positive rate increases in blind mode due to relaxed constraints; results require post-hoc filtering by FDR, manual inspection of fragmentation patterns, or orthogonal validation (e.g., genomic evidence).
• Blind mode is most effective on high-quality, well-fragmented spectra; poor spectral data or weak ionization will amplify spurious matches.
• No changelog or version-specific documentation of blind-mode algorithm changes was found; reproducibility across NPDtools versions may vary.
• Blind mode assumes the reference RiPP structure database is sufficiently comprehensive; if the true parent structure is absent, blind mode cannot recover it.

Evidence

• [methods] Execute MetaMiner again with --blind flag enabled to search for arbitrary post-translational modifications.: "Execute MetaMiner again with --blind flag enabled to search for arbitrary post-translational modifications."
• [methods] MetaMiner is a metabologenomic pipeline which integrates metabolomic (tandem mass spectra) and genomic data to identify novel RiPPs: "MetaMiner is a metabologenomic pipeline which integrates metabolomic (tandem mass spectra) and genomic data to identify novel RiPPs"
• [methods] Parse and tabulate significant_matches.tsv outputs from all three tools, recording scan identifiers, match scores, p-values, and false discovery rates.: "Parse and tabulate significant_matches.tsv outputs from all three tools, recording scan identifiers, match scores, p-values, and false discovery rates."
• [methods] Cross-compare hit sets to identify tool-specific detections, common identifications, and mode-dependent sensitivity differences.: "Cross-compare hit sets to identify tool-specific detections, common identifications, and mode-dependent sensitivity differences."
• [readme] VarQuest — a tool for modification-tolerant identification of novel variants of PNPs: "VarQuest — a tool for modification-tolerant identification of novel variants of PNPs"