Grant: Project Grant
University of Toronto
$229,500.00
Illuminating the Dark RNA Virome through ultra-deep homology search
EF 2022-23
Résumé
SARS-CoV-2 has killed millions and cost the economy trillions. To pre-empt future pandemics, we will radically increase the sensitivity of the computational tools used to detect RNA viruses. Illuminating the full diversity of Earth's viruses directly supports Canadians' health by empowering a new generation of unbiased viral diagnostics; and indirectly through better viral surveillance across all facets of One Health (monitoring of crops, wildlife, and our environments for pathogens).
At most 0.1% of Earth's viruses are known. Our ignorance is due to 1) inadequate sampling of Earth's biodiversity; and 2) poor computational sensitivity for detecting viruses. To overcome each of these limits, we propose combining two state-of-the-art technologies in a dual ultra-massive (Serratus) / ultra-sensitive (AlphaFold2) experiment.
Serratus is a cloud-computing platform allowing us to analyze the global-collection of DNA/RNA sequencing data. Collectively public sequencing data gathered over the past 15 years captures 10M+ samples (60M+ gigabytes) at a cost of $10+ billion. Previously with Serratus we analyzed 5.4 million samples and discovered >130,000 new species of RNA viruses (only 15,000 were known previously). This was based on standard methods which have less than 50% sensitivity for detecting a novel virus, those we call Dark RNA viruses.
To identify the missing 50% of Dark RNA viruses, we will deploy an AI tool called AlphaFold2 which enables vastly more sensitive protein structure analysis, albeit at ~10,000x the computational cost. Thus we "deeply" analyze 50,000 datasets to find examples of Dark RNA viruses, and in turn perform a "broad" re-analysis of all (now 8M+) public datasets.
Illuminating the RNA virome has the potential to uncover viruses associated with human diseases long believed to have an infectious causes: (re)consider Parkinson's, Crohn's, arthritis, or cancers. The goal of this project is to catalyse a new generation infectious disease research.
Programme
Project Grant
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database
Tel que publié
Id
478741
Name
Babaian Artem
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Infection and Immunity
Theme2
Biomedical
Prcname2
Genomics: Systems and computational biology
Projecttitle
Illuminating the Dark RNA Virome through ultra-deep homology search
Programtype2
Operating Grants
Orgnameinp2
University of Toronto
Cihramount2
$229,500
Cihrequipment2
$0
Programname2
Project Grant
Keyworddelim
Homology Search; Pathogen Surveillence; Petabase-Scale Sequence Search; Protein Structure; Rna Dependent Rna Polymerase (Rdrp); Rna Virus; Virus Discovery
Abstract
SARS-CoV-2 has killed millions and cost the economy trillions. To pre-empt future pandemics, we will radically increase the sensitivity of the computational tools used to detect RNA viruses. Illuminating the full diversity of Earth's viruses directly supports Canadians' health by empowering a new generation of unbiased viral diagnostics; and indirectly through better viral surveillance across all facets of One Health (monitoring of crops, wildlife, and our environments for pathogens).
At most 0.1% of Earth's viruses are known. Our ignorance is due to 1) inadequate sampling of Earth's biodiversity; and 2) poor computational sensitivity for detecting viruses. To overcome each of these limits, we propose combining two state-of-the-art technologies in a dual ultra-massive (Serratus) / ultra-sensitive (AlphaFold2) experiment.
Serratus is a cloud-computing platform allowing us to analyze the global-collection of DNA/RNA sequencing data. Collectively public sequencing data gathered over the past 15 years captures 10M+ samples (60M+ gigabytes) at a cost of $10+ billion. Previously with Serratus we analyzed 5.4 million samples and discovered >130,000 new species of RNA viruses (only 15,000 were known previously). This was based on standard methods which have less than 50% sensitivity for detecting a novel virus, those we call Dark RNA viruses.
To identify the missing 50% of Dark RNA viruses, we will deploy an AI tool called AlphaFold2 which enables vastly more sensitive protein structure analysis, albeit at ~10,000x the computational cost. Thus we "deeply" analyze 50,000 datasets to find examples of Dark RNA viruses, and in turn perform a "broad" re-analysis of all (now 8M+) public datasets.
Illuminating the RNA virome has the potential to uncover viruses associated with human diseases long believed to have an infectious causes: (re)consider Parkinson's, Crohn's, arthritis, or cancers. The goal of this project is to catalyse a new generation infectious disease research.
Competitiondate
202303
Approvedterm2
3 yrs 0 mth
Pinamesdelim
Babaian, Artem
Approvedterm
3 ans 0 mois
Cihrcontribution2
~ $229,500~ $0~