Peptide Sequencing with Deep Learning

thesis · 2020

thesis · 2020. Rui Qiao. In shotgun proteomics, de novo peptide sequencing from tandem mass spectrometry is the key technology…
Date 2020-09-30
Type thesis
Publisher PhD thesis
Contribution algorithm
Supervisor Ali Ghodsi
Link https://uwspace.uwaterloo.ca/items/58cef361-a66f-4fe2-a4aa-018da0c4762d
Citations (OpenAlex) 1

Abstract

In shotgun proteomics, de novo peptide sequencing from tandem mass spectrometry \ndata is the key technology for finding new peptide or protein sequences. It has successful applications in assembling monoclonal antibody sequences and great potentials for \nidentifying neoantigens for personalized cancer vaccines. In this thesis, I propose a novel \ndeep neural network-based de novo peptide sequencing model: PointNovo. The proposed \nPointNovo model not only outperforms the previous state-of-the-art model by a significant \nmargin but also solves the long-standing accuracy–speed/memory trade-off problem that \nexists in previous de novo peptide sequencing tools. Further, our experiment results show \nthat even though PointNovo is not trained to distinguish between true and false peptide \nspectrum matching, its resulting log probability score can be used as a scoring function \nto perform database searching. On several different datasets, we show that PointNovo, \nwhen used as a database search engine, can achieve an identification rate that is at least \ncomparable to existing popular database search softwares. \nWe also extend and adapt an existing model to process Data Independent Acquisition \n(DIA) data and propose the first de novo peptide sequencing algorithm for DIA tandem \nmass spectra. \nFinally, we develop a workflow that can identify tumor-specific antigens directly and \npurely from mass spectrometry data of tumor tissues and test it on a published dataset of \ntumor samples from melanoma patients. Our workflow applies de novo peptide sequencing \nto detect mutated endogenous peptides, in contrast to the prevalent indirect approach of \ncombining exome sequencing, somatic mutation calling, and epitope prediction in existing \nmethods. More importantly, we develop machine learning models that are tailored to each \npatient based on their own MS data. Such a personalized approach enables accurate identification of neoantigens for the development of personalized cancer vaccines. We applied \nthe workflow to datasets of five melanoma patients and expanded their immunopeptidomes \nby 5% to 15%. Subsequently, we discovered 17 neoantigens of both HLA–I and HLA–II, \nincluding those with validated T cell responses and those novel neoantigens that had not \nbeen reported in previous studies.

Authors

  1. Rui Qiao · Bioinformatics Solutions (Canada), Bioinformatics Solutions Inc., University of Waterloo

Methods and tools

Data used

  • BALF proteomics - In-depth proteomic analysis of human bronchoalveolar lavage fluid towards the biomarker discovery for (as deposited) · PXD012645
  • Diabetes causes marked inhibition of mitochondrial metabolism in pancreatic β-cells (as deposited) · PXD012979
  • High-resolution spatially-resolved proteome mapping using automated, sacrificial liquid-mediated sample transfer from la (as deposited) · PXD008844
  • Human pancreatic cancer LC-MSMS (as deposited) · PXD007890
  • Low-density lipoprotein receptor-related protein 1 (LRP1)-derived peptides protect against aggregation of LDL and choles (as deposited) · PXD011246
  • Nine-species benchmark (original (DeepNovo, 2017)) · MSV000081382, PXD003868 (provenance), PXD004325 (provenance), PXD004424 (provenance), PXD004467 (provenance), PXD004536 (provenance), PXD004565 (provenance), PXD004947 (provenance), PXD004948 (provenance), PXD005025 (provenance)
  • Proteome of the rodent malaria parasite Plasmodium berghei liver stage merosomes (as deposited) · PXD010559
  • Proteomic Analysis of Human Liver Reference Material (as deposited) · PXD009021
  • Proteomic analysis of human sclera tissue using high resolution mass spectrometry (as deposited) · PXD008999
  • Proteomic analysis of six different tissues from the Atlantic bottlenose dolphin (Tursiops truncatus) (as deposited) · PXD008808
  • Re-analysis of Hela phosphoproteome (PXD000612) with TagGraph (as deposited) · PXD008899

Cited by (6)

Seen in the charts

Back to the full map

Back to top