De novo mass spectrometry peptide sequencing with a transformer model

Abstract

Tandem mass spectrometry is the only high-throughput method for analyzing the protein content of complex biological samples and is thus the primary technology driving the growth of the field of proteomics. A key outstanding challenge in this field involves identifying the sequence of amino acids -- the peptide -- responsible for generating each observed spectrum, without making use of prior knowledge in the form of a peptide sequence database. Although various machine learning methods have been developed to address this de novo sequencing problem, challenges that arise when modeling tandem mass spectra have led to complex models that combine multiple neural networks and post-processing steps. We propose a simple yet powerful method for de novo peptide sequencing, Casanovo, that uses a transformer framework to map directly from a sequence of observed peaks (a mass spectrum) to a sequence of amino acids (a peptide). Our experiments show that Casanovo achieves state-of-the-art performance on a benchmark dataset using a standard cross-species evaluation framework which involves testing with out-of-distribution samples, i.e., spectra with never-before-seen peptide labels. Casanovo not only achieves superior performance but does so at a fraction of the model complexity and inference time required by other methods.

Prosit Transformer: A transformer for Prediction of MS2 Spectrum Intensities

“…However, in the last couple of years, a structure first in natural language processing 15 known as Transformers 16 has successfully been employed within bioinformatics, e.g., structure prediction, 17 , 18 gene expression prediction, 19 and even within MS-based proteomics, e.g., peptide detection problem, 20 DIA library generation for the phosphoproteome, 21 and de novo interpretation of MS2 spectra. 22 …”

“…However, in the last couple of years, a structure first in natural language processing 15 known as Transformers 16 has success-fully been employed within bioinformatics, e.g., structure prediction, 17,18 gene expression prediction, 19 and even within MS-based proteomics, e.g., peptide detection problem, 20 DIA library generation for the phosphoproteome, 21 and de novo interpretation of MS2 spectra. 22 Transformers are, like RNNs, designed to handle sequential input data and do so through attention mechanisms, i.e., mechanisms that enhance the essential parts of the input sequence for its output. However, unlike RNNs, the Transformers do not use recurrence, thus enabling a significant speed-up by parallelizing their training.…”

“…Many types of frameworks are available for training a predictor, such as support vector machines and recurrent neural networks (RNNs) used within MS-based proteomics. However, in the last couple of years, a structure first in natural language processing known as Transformers has successfully been employed within bioinformatics, e.g., structure prediction, , gene expression prediction, and even within MS-based proteomics, e.g., peptide detection problem, DIA library generation for the phosphoproteome, and de novo interpretation of MS2 spectra …”

Unified and standardized mass spectrometry data processing in Python using spectrum_utils

“…Cysteine carbamidomethylation was set as a fixed modification, and variable modifications were methionine oxidation, N-terminal acetylation, N-terminal carbamylation, pyroglutamate formation from glutamine, and deamidation of asparagine and glutamine. MSGF+ was configured to allow one 13 C precursor mass isotope, at most one non-tryptic terminus, and 10 ppm precursor mass tolerance. The searches were individually filtered at 1% PSM-level FDR.…”

“…This dovetails with visualization functionality to produce publication-quality and interactive spectrum graphics. Since its introduction, spectrum_utils has been used in tools that perform spectral library searching 10 and spectrum clustering, 11 to preprocess MS/MS spectra prior to deep learning applications, 12,13 to plot MS/MS spectra from online data repositories, 14 and to assist in MS/MS processing and visualization efforts for dozens of other projects. [15][16][17][18][19][20][21][22] Here we present recent updates to spectrum_utils.…”