CNL Model: HRMS Data for Fragment Detection and Model Optimization

A dataset for constructing and testing a Cumulative Neutral Loss model for fragment detection in High-Resolution Mass Spectrometry data. It contains an HRMS database, lists of database fragments and true negative fragments for model optimization, and suspect screening files for model evaluation. The dataset was contributed by Denice van Herwerden and is available under an Open Access license.

Use Cases

• Optimizing a fragment detection model based on provided lists of database fragments and true negative fragments.
• Evaluating model performance using suspect screening files that contain signals with true/false classifications.
• Analyzing the relationship between fragment signals and parent ions using provided CNL scores, Pearson correlation coefficients, and retention time differences.

Strengths

• Includes distinct components for model construction (HRMS database, fragment lists) and testing (suspect screening files).
• Provides ground truth labels (true/false classification) for signals in the test files.
• Contains multiple evaluation metrics (CNL score, Pearson correlation, retention time difference) per signal for analysis.

Limitations

• Row count, file size, and specific file formats are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Denice van Herwerden via paperswithcode

Collection Method

Likely compiled for research on the Cumulative Neutral Loss model.