Precise Discovery of Novel N-Terminal Proteoforms beyond the Limitations of Proteogenomics and De Novo Sequencing
peer-reviewed · Analytical Chemistry · 2025
| Date | 2025-08-12 |
| Type | peer-reviewed |
| Venue | Analytical Chemistry |
| Publisher | American Chemical Society (ACS) |
| Contribution | adjacent |
| DOI | 10.1021/acs.analchem.5c02498 |
| Citations (OpenAlex) | 1 |
| Venue 2-year citedness | 7.10 |
Abstract
Alternative splicing-mediated protein N-terminal sequence variation is closely associated with diseases, but its identification by mass spectrometry faces technical bottlenecks. Traditional proteogenomic methods cannot identify novel N-terminal proteins undetected in transcriptome data, while de novo sequencing has limitations in accuracy and traceability. To address this, we developed the first dedicated algorithm, NovelNSeq, which is specifically designed to parse signature peptides (novel N-terminal extension peptides) of novel N-terminal proteins from mass spectrometry data without relying on transcriptome data or de novo sequencing. NovelNSeq fully exploits peptide encoding rules, demonstrating significantly higher accuracy than de novo sequencing algorithms such as PEAKS, pNovo3, SpliceNovo, Casanovo, and InstaNovo, and enables tracing back the peptide encoding mechanisms. Using NovelNSeq, we identified and validated novel N-terminal proteoforms from human genes CALM2, CAPNS1, and CPNE7 in mass spectrometry data where large-scale proteogenomics failed to detect them, which establishes NovelNSeq as an essential complement to conventional approaches. Furthermore, we revealed that a recently reported AAG-initiated novel N-terminus in human ATP9A is actually generated through translational frameshifting from the canonical ATG start codon, highlighting the need for rigorous validation of noncanonical start codon annotations in novel N-terminal proteoforms.
Methods and tools
- NovelNSeq: Dedicated algorithm for novel N-terminal proteoforms arising from alternative splicing, parsing signature N-terminal extension peptides. Positioned as an alternative to both proteogenomics, which cannot see proteins absent from the transcriptome, and de novo sequencing, which the authors argue lacks the accuracy and traceability required here.