FREE: Enhanced Feature Representation for Isotopic Envelope Evaluation in Top-Down Mass Spectra Deconvolution

Area of Science:

• Mass spectrometry
• Proteomics
• Bioinformatics

Background:

• Deconvolution of top-down mass spectra is crucial for identifying monoisotopic peaks.
• Accurate assessment of theoretical and experimental envelope similarity is vital for mass spectra data deconvolution.
• Existing methods often lack comprehensive assessment, relying solely on intensity and m/z similarity.

Purpose of the Study:

• To develop an enhanced feature representation for isotopic envelope evaluation (FREE).
• To improve the comprehensive and refined assessment of similarity between theoretical and experimental envelopes.
• To overcome limitations of existing evaluation methods in mass spectra deconvolution.

Main Methods:

• Developed FREE to derive diverse feature representations, including peak intensity and envelope shape.
• Trained and evaluated FREE on ovarian tumor (OT) and zebrafish (ZF) datasets.
• Compared FREE's performance against state-of-the-art methods using multiple evaluation metrics.

Main Results:

• FREE demonstrated higher performance and precision across OT and ZF datasets compared to existing methods.
• FREE identified a greater number of positive envelopes and increased proteoform-spectrum matches (PrSMs) in ZF data.
• Combined FREE with TopFD identified more fragment ions and PrSMs across diverse proteomes, outperforming EnvCNN.

Conclusions:

• FREE provides a robust and generalized approach for isotopic envelope evaluation in top-down mass spectrometry.
• The enhanced feature representation significantly improves the accuracy and comprehensiveness of mass spectra deconvolution.
• FREE facilitates more accurate identification of proteoforms and fragment ions across various biological samples.