Speaker
Description
Most top-down proteomics workflows rely on deconvolution of intact and fragment ion m/z values using modeled isotope distributions, typically via an “averagine” approximation. This step often limits accuracy: poor fits to distorted isotope patterns can lead to incorrect monoisotopic mass assignment, widened mass tolerances, and inflated false discovery rates. To address these limitations, we have developed a framework for de novo sequencing and internal calibration that operates entirely in natural log-transformed m/z space—eliminating the need for monoisotopic mass determination.
By transforming spectra to ln(m/z − q), where q is the charge carrier mass, peaks arising from the same analyte mass align along a predictable pattern defined solely by charge state—a principle formalized by Jeong et al. in the FLASHDeconv algorithm (2020). This mass-invariant spacing can be used to assign charge states, pair isotopologues, and perform internal calibration without averagine-based fitting. Calibration is achieved by optimizing the B coefficient in the Ledford equation until observed peaks align with the expected −ln(c) spacing. Sequence tag inference is performed by comparing log-transformed peak positions from consecutive fragment ions to expected values based on known residue mass differences. When observed ln(m/z − q) values match those predicted for a given residue across multiple isotopologues and charge states, the corresponding mass difference can be confidently assigned—even from a single scan.
This method was applied to 21 T FT-ICR MS/MS spectra of intact proteins, achieving sub-ppm agreement between predicted and observed values without spectral averaging. Internal calibration improved mass accuracy of myoglobin from 6.9 ppm RMSE to 0.8 ppm. Notably, near-isobaric residues such as lysine and glutamine were resolved at high charge state, and proteoform families were identified from MS¹ data using log-space mass differences alone. This database-independent, calibrant-free framework enables high-accuracy proteoform analysis and significantly improves the robustness and resolution of top-down de novo sequencing.
| User consent | yes |
|---|
