Machine Learning and Big Data as Tools for Chemistry Research

Safeguarding food integrity is about ensuring the safety, quality, and authenticity of food products. We, the FACT Lab, focus on the advancement and application of analytical chemistry and technologies to solve the persisting and emerging food integrity issues faced by the food industry. We take an interdisciplinary approach by organically integrating Analytical Chemistry, Nanomaterials, Microfluidic “Lab-on-a-Chip” Platform, Metabolomics, Bioinformatics, and Data Science to develop simple, rapid, cost-effective, reliable and smart point-of-need sensors and instrumental methods to monitor and trace food products.

Using these novel technologies, we aim to provide the last barrier of protection to food integrity from farm to fork and thus to protect consumers, genuine food industries and government, as well as to reduce food waste and loss during the supply chain.

For more information, visit The FACT Lab

The Rowley Group uses computer simulations to study biological and materials chemistry. We use a combination of molecular simulation, quantum chemistry, and machine learning to understand complex chemical systems and then to use these ideas to predict and design new chemical systems. Our core areas of interest are protein–drug binding, covalent modifier drugs, membrane permeation, and cryo-EM structure refinement. To make our models more accurate and to explore new types of chemistry, we develop new methods for computational chemistry, such as our hybrid neural network potential / molecular mechanical model.

For more information, visit the Rowley Group.

Overview

Mass spectrometry (MS) has proven to be a powerful analytical tool in many scientific fields for nearly a century. Advances made in soft ionization techniques such as electrospray ionization (ESI) and matrix-assisted laser desorption ionization (MALDI) have expanded the use of MS to the biological sciences, allowing biomolecules to be ionized and sampled in the gas phase without fragmentation. These advances have been echoed in an explosion of publications over the past 20 years that have used MS as a tool to illuminate information pertaining to a biological question. The majority of MS-based bioanalytical studies to date have focused on cataloguing the species present in a sample at a static moment in time. Although this is important, measuring the dynamics of a system in response to a stimulus not only aids identification efforts, but offers mechanistic insight into the cellular processes involved. Gaining a better understanding of these dynamics, often expressed through a change in concentration or in the degree of modification, enhances the accuracy of cellular models, assigns functional roles to the identified species, as well as highlights new diagnostic and therapeutic strategies. The research program of the Smith group focuses on developing novel methods to efficiently measure biomolecular dynamics in living systems with the downstream goals of elucidating disease biomarkers and further characterizing biosynthetic and biochemical pathways.

Instrumentation

The Smith Lab uses mass spectrometry as a tool to probe the dynamics of various biomolecules. At present, we mainly conduct research in the area of protein analysis (“proteomics”) and lipid analysis (“lipidomics”). The laboratory is equipped with high-resolution hybrid quadrupole time of flight, triple quadrupole and linear ion trap mass spectrometers. All of these instruments achieve ionization via electrospray ionization and are coupled to high performance liquid chromatographs. The lab additionally houses two gas chromatography-MS systems, one equipped with a headspace sampler.

TrEnDi

Trimethylation Enhancement using Diazomethane (TrEnDi) is a published technique developed by Professors Jeffrey Smith and Jeffrey Manthorpe at Carleton University. The method increases the sensitivity of mass spectrometric detection by assigning a fixed, permanent positive charge to amino groups. It allows for increased and predictable sequence coverage for peptides in proteomic analyses, and increased limits of detection for several important lipid classes in lipidomic analyses.

Wasslen, K.V., Canez, C.R., Lee, H., Manthorpe, J.M., Smith, J.C. (2014) Trimethylation Enhancement using Diazomethane (TrEnDi) II: Rapid In-Solution Concomitant Quaternization of Glycerophospholipid Amino Groups and Methylation of Phosphate Groups via Reaction with Diazomethane Significantly Enhances Sensitivity in Mass Spectrometry Analyses via a Fixed, Permanent Positive Charge. Analytical Chemistry. Oct 7;86(19): 9523-9532. DOI: 10.1021/ac501588y. PMID: 25208053.

Quantitative Proteomics

The Smith Lab uses microfluidic devices to manipulate biological samples in a highly efficient manner to elucidate protein dynamics and post-translational modification patterns using MS in conjunction with TrEnDi. Through studying the dynamics of protein abundances or the modes in which they are modified in cells that are stimulated in some manner (or that are either diseased or healthy), our research will help define how cells communicate with each other, their environment and themselves.

Quantitative Lipidomics

Although lipids have been studied for decades and have been largely regarded as energy storage or structural molecules, recent progress in lipid research has revealed many novel and important roles for lipids in cellular signaling. Many classes of lipids are easily analyzed by mass spectrometry; however, some are more difficult to observe. The Smith Laboratory will focus on developing novel methods to identify lipid species in the context of complicated biological samples. To date, the development of TrEnDi has achieved this goal by increasing the sensitivity of some lipid classes over an order of magnitude. Ultimately, we will apply the novel methods that are developed to investigate lipid dynamics in biological systems to aid our understanding of the roles they play in cellular life.

For more information, visit The Smith Lab.

The Barry Lab is interested in the development of precursors and processes for atomic layer deposition (ALD) and was the first academic research group in Canada to work in this field.

We have previously discovered processes for the deposition of the coinage metals (Cu, Au), used in microelectronic chip manufacturing and sensing applications. Our present research is centred on earlier transition metals (Co, Ni, Mo, W) for next-generation nano-electronic interconnects, as well as upgrading additive manufacturing processes through thin film coating and sequential infiltration.

As synthetic chemists, we try to determine the mechanism of the surface chemistry and thermal decomposition routes to better design precursors and tune our processes. We look at many different processes and target films, but the theme of mechanistic inorganic chemistry can broadly be drawn through our work.

For more information, visit the Barry Lab.