Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Doctoral Research Award - PA - Bioinformatics and Computational Biology

Direct Payment

$120,000.00

Image-based representation learning for improved genome-wide association discovery in human retinal, retinovascular, and cardiovascular systems.

EF 2024-25

Résumé
Our genetics informs a great deal about how our bodies are designed. While some of our traits or diseases (termed 'phenotype') may be controlled by a single gene, almost all human phenotypes are the result of the combined effect of many genes across our individual chromosomes. Genome-wide association studies (GWAS) are a statistical approach to identify the genes linked to a particular phenotype: genomic information is gathered from a large sample of the population, who will exhibit the trait of interest in varying degrees; the analysis reveals whether the trait is significantly correlated with certain genetic variations. Over the past 20 years, GWAS has helped shed light on the genetic basis for many traits (eg. height, BMI) and diseases (eg. diabetes, multiple sclerosis). However, while some traits such as height or a positive diagnosis are easy to measure or label, more complex traits, such as the structure of the heart, are difficult to summarise in a way that is suitable for genomic analysis. Consequently, GWAS on such traits thus far have relied on proxy measures, such as ventricular volume to study heart structure, but these do not adequately capture the complexity of our anatomy. This project aims to use the latest advances in AI to learn these complex features from medical image data to determine novel genes associated with our anatomies. The focus will be on physiological systems for which GWAS have been conducted in the past (thus some genetic variants are already known), but used limited proxy measures: the eye (retinal), heart (cardiovascular), and eye circulation (retinovascular) systems. Not only are these structures rich in spatial information and thus well-suited to image-based analysis, they are also implicated in many diseases. With better understanding of their genetics, we can improve disease diagnosis from genetic information, increase our knowledge of genetic predispositions to eye- and heart-based conditions, and find new therapeutic targets.
Programme
Doctoral Research Award - PA - Bioinformatics and Computational Biology
Bénéficiaire
Emplacement
Germany
Type
Grant
Source
CIHR Funding Decisions Database

Tel que publié

Id
517751
Name
Yoo Esther
Country
Germany
Orgtype
Research Institution
Instname2
Genetics
Theme2
Biomedical
Prcname2
Doctoral Research Awards - A
Projecttitle
Image-based representation learning for improved genome-wide association discovery in human retinal, retinovascular, and cardiovascular systems.
Programtype2
Studentship Programs
Orgnameinp2
Direct Payment
Cihramount2
$120,000
Cihrequipment2
$0
Programname2
Doctoral Research Award - PA - Bioinformatics and Computational Biology
Keyworddelim
Bioinformatics; Cardiovascular System; Computer Vision; Genome-Wide Association Study (Gwas); Genomics; Human Genetics; Machine Learning; Medical Imaging; Retinal System; Retinovascular System
Abstract
Our genetics informs a great deal about how our bodies are designed. While some of our traits or diseases (termed 'phenotype') may be controlled by a single gene, almost all human phenotypes are the result of the combined effect of many genes across our individual chromosomes. Genome-wide association studies (GWAS) are a statistical approach to identify the genes linked to a particular phenotype: genomic information is gathered from a large sample of the population, who will exhibit the trait of interest in varying degrees; the analysis reveals whether the trait is significantly correlated with certain genetic variations. Over the past 20 years, GWAS has helped shed light on the genetic basis for many traits (eg. height, BMI) and diseases (eg. diabetes, multiple sclerosis). However, while some traits such as height or a positive diagnosis are easy to measure or label, more complex traits, such as the structure of the heart, are difficult to summarise in a way that is suitable for genomic analysis. Consequently, GWAS on such traits thus far have relied on proxy measures, such as ventricular volume to study heart structure, but these do not adequately capture the complexity of our anatomy. This project aims to use the latest advances in AI to learn these complex features from medical image data to determine novel genes associated with our anatomies. The focus will be on physiological systems for which GWAS have been conducted in the past (thus some genetic variants are already known), but used limited proxy measures: the eye (retinal), heart (cardiovascular), and eye circulation (retinovascular) systems. Not only are these structures rich in spatial information and thus well-suited to image-based analysis, they are also implicated in many diseases. With better understanding of their genetics, we can improve disease diagnosis from genetic information, increase our knowledge of genetic predispositions to eye- and heart-based conditions, and find new therapeutic targets.
Competitiondate
202410
Approvedterm2
3 yrs 0 mth
Pinamesdelim
Yoo, Esther
Approvedterm
3 ans 0 mois
Cihrcontribution2
~ $120,000~ $0~