Grant: Discovery Launch Supplement
Dalhousie University
$12,500.00
Novel Strategies for the Analysis of Multidimensional Data in the Omics Sciences
EF 2024-25
Résumé
Research in Molecular Biology has recently become reliant on large amounts of data with complex structures that hide invaluable information. This proposal describes my research program which aims to develop new computational tools that are best suited for extracting the maximum amount of information from multidimensional data that now pervade Functional Genomics (FG). FG aims to find relationships between an organism's genotype (genes, etc.) and its current phenotype using bio-analytical technologies and is different from Comparative Genomics (CG) where, e.g., de novo sequence assembly, sequence comparisons etc., are sought. While the more mature CG is supported by an extended array of bioinformatics tools, FG still faces unique challenges because the rate of development of bio-analytical technologies outpaces that of computational methods to deal with the data.
There are important challenges facing FG research in its present form. First, the data generated are typically very poorly characterised, and a clutch of computational methods are "thrown at them" because of the need to extract some biological insight. These methods are often borrowed from other fields and employed regardless of inherent data characteristics. Second, for most biologists in academic labs, these experiments are very not only very costly, they generate a deluge of data that require computational expertise that is critically needed. To deal with this, experiments are conducted in a way that can answer the simplest questions. For certain studies, it is inevitable that more complex experimental protocols are pursued, producing extremely complex data. Finally, although the so-called "multi-omics" data are now routinely generated, methods to analyze them are severely lacking, creating an existential bottleneck.
Although FG data are continuously evolving, they exhibit unique characteristics including large differences (orders of magnitude) in their linear dynamic range, non-uniform measurement errors (MEs), nonlinear relationships and auto-correlations among the variables, especially for temporal (ordinal) experiments. These issues particulary impede access to the information embedded in the low intensity signals. My program will continuously develop new data analysis approaches that focus on the core structures of the measurement systems and provide optimal solutions to access information embedded in these low intensity signals. In the initial 5-year phase, we will (a) characterize the structure of measurement errors (MEs) associated with label-free quantification by liquid chromatography coupled to mass spectrometry (LFQ-LC-MS), and incorporate these MEs in modeling the data (b) develop novel strategies to model data derived from the more complex, time-ordered (ordinal) measurements, and (c) introduce projection pursuit as a non-canonical alternative approach for subspace estimation which will allow data derived from multiple bioanalytical platforms to be integrated.
Programme
Discovery Launch Supplement
Bénéficiaire
Emplacement
Nova Scotia, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
DGECR-2024-00132
Name nom
Karakach, Tobias
Department département
Pharmacology
Organization id
79
Institution établissement
Dalhousie University
Province en
Nova Scotia
Province fr
Nouvelle-Écosse
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
12500
Program id
DGECR
Program name en
Discovery Launch Supplement
Program name fr
Supplément Tremplin vers la découverte
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1501
Committee name en
Genes, Cells and Molecules
Committee name fr
Gènes, cellules et molécules
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
5300
Research subject group en
Molecular biology
Research subject group fr
Biologie moléculaire
Research subject en
Molecular biology
Research subject fr
Biologie moléculaire
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
Novel Strategies for the Analysis of Multidimensional Data in the Omics Sciences
Keywords
proteomics; mass spectrometry; bioinformatics; measurement error models; metabolomics; pattern recognition; time course measurements; hidden markov models; projection pursuit; low intensity signals
Application summary
Research in Molecular Biology has recently become reliant on large amounts of data with complex structures that hide invaluable information. This proposal describes my research program which aims to develop new computational tools that are best suited for extracting the maximum amount of information from multidimensional data that now pervade Functional Genomics (FG). FG aims to find relationships between an organism's genotype (genes, etc.) and its current phenotype using bio-analytical technologies and is different from Comparative Genomics (CG) where, e.g., de novo sequence assembly, sequence comparisons etc., are sought. While the more mature CG is supported by an extended array of bioinformatics tools, FG still faces unique challenges because the rate of development of bio-analytical technologies outpaces that of computational methods to deal with the data.
There are important challenges facing FG research in its present form. First, the data generated are typically very poorly characterised, and a clutch of computational methods are "thrown at them" because of the need to extract some biological insight. These methods are often borrowed from other fields and employed regardless of inherent data characteristics. Second, for most biologists in academic labs, these experiments are very not only very costly, they generate a deluge of data that require computational expertise that is critically needed. To deal with this, experiments are conducted in a way that can answer the simplest questions. For certain studies, it is inevitable that more complex experimental protocols are pursued, producing extremely complex data. Finally, although the so-called "multi-omics" data are now routinely generated, methods to analyze them are severely lacking, creating an existential bottleneck.
Although FG data are continuously evolving, they exhibit unique characteristics including large differences (orders of magnitude) in their linear dynamic range, non-uniform measurement errors (MEs), nonlinear relationships and auto-correlations among the variables, especially for temporal (ordinal) experiments. These issues particulary impede access to the information embedded in the low intensity signals. My program will continuously develop new data analysis approaches that focus on the core structures of the measurement systems and provide optimal solutions to access information embedded in these low intensity signals. In the initial 5-year phase, we will (a) characterize the structure of measurement errors (MEs) associated with label-free quantification by liquid chromatography coupled to mass spectrometry (LFQ-LC-MS), and incorporate these MEs in modeling the data (b) develop novel strategies to model data derived from the more complex, time-ordered (ordinal) measurements, and (c) introduce projection pursuit as a non-canonical alternative approach for subspace estimation which will allow data derived from multiple bioanalytical platforms to be integrated.