In Silico Fragment Prediction

in-silico-fragment-prediction

Summary

Generate predicted mass spectrometry fragment spectra computationally from compound chemical structures using CFM-ID, enabling rapid annotation and validation of experimental fragmentations for metabolites and DNA adducts without requiring wet-lab MS/MS experiments.

When to use

Apply this skill when you have a collection of compound structures in SDF format (e.g., DNA adduct structures) and need to systematically generate predicted fragment spectra across a defined ionization level and mass range to populate a reference spectral database or validate experimental fragmentation patterns.

When NOT to use

• Input compounds lack well-defined 2D/3D structure information — CFM-ID requires chemically valid structure representations
• Experimental fragment spectra are already available and validated — in-silico prediction is most useful for annotation of unknowns or high-throughput reference generation, not as a replacement for experimental data
• The goal is to identify which compounds are present in a sample — use this skill for spectral reference building, not for sample analysis

Inputs

• SDF format compound structure file containing molecular structures (e.g., DNA adduct compounds)

Outputs

• Structured predicted fragment spectra database indexed by compound identifier
• Individual predicted MS/MS spectra for each input compound

How to apply

Load all compound structures from an SDF format file containing your target compounds. Execute CFM-ID on each structure, specifying the appropriate ionization level and mass range relevant to your compounds. Compile the resulting predicted fragment spectra into a structured database matching the schema of your target resource (e.g., matching the format of online predicted-fragments databases). Validate completeness by confirming that all input compounds from the SDF file have corresponding predicted spectra entries in the output database.

Related tools

• CFM-ID (In-silico fragment spectrum prediction engine that processes compound structures and generates predicted MS/MS spectra at specified ionization levels)

Evaluation signals

• All compounds in the input SDF file have exactly one corresponding entry in the output predicted fragments database
• Predicted spectra follow expected mass fragmentation patterns consistent with the chemical structure and ionization method (no negative masses, fragments within specified mass range)
• Output database schema matches the format of the target resource (e.g., predicted-fragments online database format) with required fields populated
• Predicted spectra can be successfully cross-matched against experimental spectra using standard metrics (e.g., cosine similarity with experimental MS/MS data when available)

Limitations

• CFM-ID prediction accuracy depends on compound structure quality and chemical validity; invalid or poorly-drawn structures produce unreliable spectra
• Predicted spectra represent gas-phase fragmentation patterns and may not fully capture solution-phase or matrix-dependent effects observed in experimental data
• The skill requires specification of correct ionization level and mass range parameters; misaligned parameters will generate spectra outside the relevant analytical window

Evidence

• [other] Load compound structures from the SDF format file containing DNA adduct compounds: "Load compound structures from the SDF format file containing DNA adduct compounds"
• [other] Execute CFM-ID on each compound structure to predict fragment spectra at the appropriate ionization level and mass range: "Execute CFM-ID on each compound structure to predict fragment spectra at the appropriate ionization level and mass range"
• [other] Compile predicted fragment spectra into a structured database matching the format of the deposited predicted-fragments online resource: "Compile predicted fragment spectra into a structured database matching the format of the deposited predicted-fragments online resource"
• [other] Validate output by confirming all input compounds have corresponding predicted spectra entries: "Validate output by confirming all input compounds have corresponding predicted spectra entries"
• [other] The in-silico fragment prediction stage uses CFM-ID to process SDF compound structures and generate predicted fragment spectra: "The in-silico fragment prediction stage uses CFM-ID to process SDF compound structures and generate predicted fragment spectra"