Shotgun proteomics based on high-resolution mass spectrometry (MS) is transforming biological research in the post-genomic era and offers an unbiased analytical profiling method that complements existing targeted MS and antibody-based approaches. However, using current technology only about 15-30% of the >100,000 observable individual peptide isotopic patterns¹ are ever identified.  During traditional data-dependant MS acquisition (DDA) of complex samples, nearly all MS/MS spectra generated contain contaminating fragment ions that arise from the isolation and fragmentation of multiple co-eluting peptides. Identification of these chimeric spectra is limited by the ‘one MS/MS spectrum–one peptide’ strategy used in nearly all available MS/MS search algorithms. 

Here we report MSCypher, a new proteomics workflow that combines an improved tandem MS acquisition strategy with a novel search algorithm. Our initial benchmarking of MSCypher shows significant improvements in speed and sensitivity of peptide detection and identification. This high-resolution MS approach takes advantage of co-eluting peptides and uses a dynamic isolation window to evenly distribute the number of co-fragmented species during acquisition. Processing these multiplexed spectra datasets has required a redesign of the current analysis workflow to include retention time prediction, dynamic probabilistic fragment ion matching and a supervised ensemble learning method for peptide feature identification.

Our initial results indicate that this multiplexed MS/MS approach provides both enhanced sensitivity and a gain in coverage. Each of these improvements was proportional to the number of peptides simultaneously analysed. We also maintain high confidence levels for peptides identified and our latest experiments indicate that identification of modified peptides will also be improved using this strategy.