Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Discovery Grants Program - Individual

Simon Fraser University

$41,000.00

Decoding the genotype-phenotype relationship through interpretable, robust and integrative machine learning

FY 2023-24

Summary
A fundamental question of biology concerns the relationship between an organism's DNA sequence (its genotype) and its observable traits (its phenotype). Solving this genotype-phenotype problem will enable understanding and engineering organisms towards a wide variety of goals including biotechnology, agriculture and health. Towards this end, researchers have recently sequenced of thousands of genomes and collected of tens of thousands of genome-wide measurements, providing a unique opportunity to address this genotype-phenotype problem. Most consequential genetic variation in mammals causes changes in gene regulation. Gene regulation is the process by which a cell determines which genes to express as proteins. This process allows for an organism's single genome to express hundreds of tissues and cell types. Thus, understanding gene regulation is key to understanding genetic differences between individuals. A complete understanding of gene regulation promises to lead to monumental advances in bioengineering and health. Yet many aspects of gene regulation remain poorly understood. The recent development of sequencing-based genomic assays promises to grant insight into gene regulation. These assays use high-throughput DNA sequencing as a readout to measure genomic activities including transcription, DNA-protein binding, chromatin accessibility and others. As of this writing, international consortia such as CEEHRC, ENCODE and IHEC have performed tens of thousands of genomic assays in a variety of tissues from organisms including human, mouse and fruit fly. In this project, we will address the genotype-phenotype problem by developing methods in the field of machine learning. Machine learning is the field that aims to develop computer algorithms that improve with experience. The rapid development of machine learning methods provide a means to tackle the challenges posed by the scale and complexity of genomic data sets. My research program is devoted to addressing the genotype-phenotype problem through accurate and interpretable machine learning models of genomic data sets. Specifically, we have three aims. First, we aim to identify genomic regulatory elements through probabilistic graphical models. Second, we aim to understand genetic variation by interpreting neural network models. Third, we aim to measure confidence in genomic predictions according to their empirical reproducibility. This research will significantly expand our understanding of genomic function, will result in machine learning methods applicable across a wide variety of fields, and will improve equity and transparency in genomic research.
Program
Discovery Grants Program - Individual
Location
British Columbia, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
RGPIN-2023-04184
Name nom
Libbrecht, Maxwell
Department département
Computing Science
Organization id
5
Institution établissement
Simon Fraser University
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
41000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1201
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Life sciences (including biotechnology)
Area of application fr
Sciences de la vie (y compris biotechnologie)
Research subject code
5004
Research subject group en
Genetics
Research subject group fr
Genome analysis
Research subject en
Génétique
Research subject fr
Analyse du génome
Application title
Decoding the genotype-phenotype relationship through interpretable, robust and integrative machine learning
Keyword
Chromatin architecture; Chromatin state; Deep learning; Epigenomics; Functional genomics; Gene regulation; Genomics; Interpretable machine learning; Probabilistic graphical models; Representation learning
Application summary
A fundamental question of biology concerns the relationship between an organism's DNA sequence (its genotype) and its observable traits (its phenotype). Solving this genotype-phenotype problem will enable understanding and engineering organisms towards a wide variety of goals including biotechnology, agriculture and health. Towards this end, researchers have recently sequenced of thousands of genomes and collected of tens of thousands of genome-wide measurements, providing a unique opportunity to address this genotype-phenotype problem. Most consequential genetic variation in mammals causes changes in gene regulation. Gene regulation is the process by which a cell determines which genes to express as proteins. This process allows for an organism's single genome to express hundreds of tissues and cell types. Thus, understanding gene regulation is key to understanding genetic differences between individuals. A complete understanding of gene regulation promises to lead to monumental advances in bioengineering and health. Yet many aspects of gene regulation remain poorly understood. The recent development of sequencing-based genomic assays promises to grant insight into gene regulation. These assays use high-throughput DNA sequencing as a readout to measure genomic activities including transcription, DNA-protein binding, chromatin accessibility and others. As of this writing, international consortia such as CEEHRC, ENCODE and IHEC have performed tens of thousands of genomic assays in a variety of tissues from organisms including human, mouse and fruit fly. In this project, we will address the genotype-phenotype problem by developing methods in the field of machine learning. Machine learning is the field that aims to develop computer algorithms that improve with experience. The rapid development of machine learning methods provide a means to tackle the challenges posed by the scale and complexity of genomic data sets. My research program is devoted to addressing the genotype-phenotype problem through accurate and interpretable machine learning models of genomic data sets. Specifically, we have three aims. First, we aim to identify genomic regulatory elements through probabilistic graphical models. Second, we aim to understand genetic variation by interpreting neural network models. Third, we aim to measure confidence in genomic predictions according to their empirical reproducibility. This research will significantly expand our understanding of genomic function, will result in machine learning methods applicable across a wide variety of fields, and will improve equity and transparency in genomic research.