Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Discovery Grants Program - Individual

University of Alberta

$40,000.00

Acceleration of Genomic Data Processing Using Derived References, Heterogeneous Parallel Computing, Computational Memory, and Low-Latency Networked Mass Storage

FY 2024-25

Summary
The objective is to develop faster and more economical genetic data processing systems that exploit new algorithms, parallel computing, fast networked memory & custom hardware accelerators to improve the way in which genetic information is recovered from DNA sequencers, a key bottleneck in genomics. The proposed program focuses first on accelerating a new class of deoxyribonucleic acid (DNA) alignment algorithms that exploit large index tables that are computed from references that are derived from genomes. Sequenced DNA read fragments are streamed from user workstations to networked accelerators co-located with the tables. The program will provide timely & marketable training for at least 5 graduate students. Over the past 30 years dramatic reductions in the cost of sequencing DNA combined with advances in bioinformatics (computer processing of biological data) have led to a Bioinformatics Revolution (BR). The genetic blueprint of an organism is its genome, which (apart from RNA viruses) is encoded in long DNA molecules, which are sequences of pairs of nucleotides of four base types: A, C, G & T. One human genome consists of ~3.2 billion base pairs. DNA sequencers rapidly determine the base sequence (called a read) of the first ~150 bases at the ends of 100s of millions of fragments of DNA that have been obtained from a sample. Computer processing then reconstructs the specific genome from the sequenced reads. Typically the reconstruction uses as a scaffold a known reference genome for the organism's species, and each read is aligned to the best matching positions in the reference, taking into account any mutations that distinguish the organism's genome from the reference. The mutations include nucleotide substitutions, insertions, deletions and more complex structural changes in the DNA. The mutations in a genome are required by bioscientists to understand the mechanisms that underlie healthy and diseased cell function, to maximize agricultural productivity, to develop more effective personalized medical treatments, to detect the presence of cancerous cells, etc. The vast potential of the BR has motivated large-scale projects where millions of genomes are being sequenced, sometimes multiple times using different references. Read alignment algorithms already use indexes that speed up the finding of matches. The FM index stores the match positions in the reference of all suffixes of all subsegments of the reference. Using this index vastly reduces the alignment time. The time can be reduced further using a jump index that lists all possible match locations for a fixed initial seed, say 15 bases. The seed match is then extended in steps of 1 or 2 bases with an FM index. Matches that require mutations (substitutions, insertions, deletions) are faster to find using novel derived references. New algorithms exploit multiple jumps into indexes based on these references, as well as hardware accelerators, to significantly speed up read alignment.
Program
Discovery Grants Program - Individual
Location
Alberta, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
RGPIN-2024-05836
Name nom
Cockburn, Bruce
Department département
Department of Electrical and Computer Engineering
Organization id
9
Institution établissement
University of Alberta
Province en
Alberta
Province fr
Alberta
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
40000
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
1510
Committee name en
Electrical and Computer Engineering
Committee name fr
Génie électrique et informatique
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
2704
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Distributed and parallel processing
Research subject fr
Traitement réparti et simultané
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Acceleration of Genomic Data Processing Using Derived References, Heterogeneous Parallel Computing, Computational Memory, and Low-Latency Networked Mass Storage
Keywords
dna sequence alignment; heterogeneous parallel computing hardware; genetic data processing; hardware accelerators for bioinformatics; field-programmable gate array (fpga) applications; hardware-software co-design; networked computational memory; cloud and edge computing; disaggregated mass storage; high-level hardware synthesis applications
Application summary
The objective is to develop faster and more economical genetic data processing systems that exploit new algorithms, parallel computing, fast networked memory & custom hardware accelerators to improve the way in which genetic information is recovered from DNA sequencers, a key bottleneck in genomics. The proposed program focuses first on accelerating a new class of deoxyribonucleic acid (DNA) alignment algorithms that exploit large index tables that are computed from references that are derived from genomes. Sequenced DNA read fragments are streamed from user workstations to networked accelerators co-located with the tables. The program will provide timely & marketable training for at least 5 graduate students. Over the past 30 years dramatic reductions in the cost of sequencing DNA combined with advances in bioinformatics (computer processing of biological data) have led to a Bioinformatics Revolution (BR). The genetic blueprint of an organism is its genome, which (apart from RNA viruses) is encoded in long DNA molecules, which are sequences of pairs of nucleotides of four base types: A, C, G & T. One human genome consists of ~3.2 billion base pairs. DNA sequencers rapidly determine the base sequence (called a read) of the first ~150 bases at the ends of 100s of millions of fragments of DNA that have been obtained from a sample. Computer processing then reconstructs the specific genome from the sequenced reads. Typically the reconstruction uses as a scaffold a known reference genome for the organism's species, and each read is aligned to the best matching positions in the reference, taking into account any mutations that distinguish the organism's genome from the reference. The mutations include nucleotide substitutions, insertions, deletions and more complex structural changes in the DNA. The mutations in a genome are required by bioscientists to understand the mechanisms that underlie healthy and diseased cell function, to maximize agricultural productivity, to develop more effective personalized medical treatments, to detect the presence of cancerous cells, etc. The vast potential of the BR has motivated large-scale projects where millions of genomes are being sequenced, sometimes multiple times using different references. Read alignment algorithms already use indexes that speed up the finding of matches. The FM index stores the match positions in the reference of all suffixes of all subsegments of the reference. Using this index vastly reduces the alignment time. The time can be reduced further using a jump index that lists all possible match locations for a fixed initial seed, say 15 bases. The seed match is then extended in steps of 1 or 2 bases with an FM index. Matches that require mutations (substitutions, insertions, deletions) are faster to find using novel derived references. New algorithms exploit multiple jumps into indexes based on these references, as well as hardware accelerators, to significantly speed up read alignment.