Oral Presentation 25th Annual Lorne Proteomics Symposium 2020

 Phosphomatics: A knowledge-based approach to investigating high-throughput phosphoproteomics data (#50)

Michael G Leeming 1 , Sean O’Callaghan 2 , Ching-Seng Ang 1 , Shuai Nie 1 , Swati Varshney 1 , Syeda Sadia Ameen 1 , Heung-Chin Cheng 1 , Nicholas A Williamson 1
  1. Bio21 Mass Spectrometry and Proteomics Facility, University of Melbourne, Parkville, VIC, Australia
  2. Nuritas Ltd, Dublin, Ireland

Mass spectrometry-based phosphoproteomics is undoubtedly one of the most powerful tools available for investigating the detailed molecular events that occur in response to cellular stimulus and, while these experiments can routinely detect and quantify thousands of phosphorylated peptides, interpreting this data remains challenging. Identification of the upstream kinases that affect the observed phosphorylations is key to understanding the factors responsible for differences in observed phosphorylation profiles however the astounding complexity of substrate-kinase relationships make this task difficult. For example, protein kinases frequently have broad but overlapping sets of substrates and protein function is often differentially regulated by selective phosphorylation at different sites. Making the most of our phosphoproteomics data clearly requires development of a knowledge base against which phosphoproteome data can be easily queried.

Here, we present ‘Phosphomatics’ – a new web-based tool for interrogating possible upstream kinases for observed phosphopeptides. Phosphomatics allows users to upload the results of global phosphoproteomics experiments containing thousands of phosphorylated peptides and search each of these against databases of known substrate-kinase relationships. Users can then interactively explore substrate-kinase relationships constructed from their input data based upon low-throughput, manually curated literature sources and results can be interrogated from either substrate-centric, kinase-centric or pathway-centric perspectives. For specific substrate-kinase relationships, machine learning based tools, in addition to publications drawn from existing databases, are utilised to provide users as much evidence as possible to support the selected phosphorylation event. Users can download results files summarising possible substrate-kinase relationships as well as graphics providing a visual overview of phosphorylation networks and important kinases.

Phosphomatics is freely available via the internet at: www.phosphomatics.com