Below is a list of abstracts by fourth-year students in the Integrated Science Program:


Dusan Vukmirovic – The effect of low dose ionizing radiation and UV radiation on mammalian.

Supervisor: Carmel Mothersill
The effects of ionizing radiation on biota have primarily been associated with direct exposure to the source. However, studies in the field suggest that damage, mainly in the form of cell killing, can also occur through signalling molecules communicated from irradiated cells to non-irradiated cells. The term ‘radiation induced bystander effect’, (RIBE), describes this phenomenon as a ‘non-targeted effect’ of radiation exposure. In addition to the RIBE that occur at the cellular level, RIBE have also been documented to occur between tissues and even at the organismal level. Of particular interest are the effects of intraorganismal and interorganismal bystander effects stemming from radiation treatments. This thesis project examines RIBE with respect to a novel form of radiotherapy, Synchrotron Microbeam Radiation Therapy (Synchrotron MRT) in the treatment of glioma blastfoma, a rare form of brain cancer. The methodology includes the application of fluorescent microscopy in calcium pulse detection, a characteristic sign of the bystander signal, with the use of a HaCaT reporter cell line. The sample organisms comprise of groups of tumour free and tumour bearing Fisher rats exposed to different doses of Synchrotron MRT and their corresponding non-irradiated cagemates. The samples contain media from tissue explants in the lateral and contralateral cerebral hemispheres and the bladder in each of these groups. The implications of this research might have profound impact on the viability of Synchrotron MRT as an effective radiotherapy treatment.

Angela Huynh – Characterization od extracellular fragments of MARCO.

Supervisor: Alba Guarné
Streptococcus pneumoniae
 is one of the leading causes of bacterial pneumonia, meningitis, and sepsis in children. Major steps have been taken to combat this pathogen. An especially effective method is the administration of the pneumococcal vaccine. Although this vaccine does protect against a variety of S. pneumoniae strains responsible for disease, the vaccine is still ineffective against many strains. Another drawback of the available pneumococcal vaccines is poor success in young children who are the most burdened by pneumococcal diseases. A booster vaccine against S. pneumoniae is needed to reduce the number of patients affected. In order to discover a novel vaccine target for S. pneumoniae, we need to understand how it is recognized and eliminated in the body. By determining the specific target which induces the immune response for all strains, determining, a vaccine target can be chosen. Macrophage receptor with collagenous region (MARCO) has a binding site within its scavenger receptor cysteine rich (SRCR) region that is believed to be responsible for recognizing S. pneumoniae. However, what MARCO recognizes on S. pneumoniae has not yet been identified. We will develop an experimental system to identify possible targets in S. pneumoniae that MARCO recognizes. An X-ray crystal structure will be determined to get a better idea of what contacts within the binding region are used to specify ligands. In addition, a number of assays will be run to isolate possible ligands of S. pneumoniae. Through the determination of a novel ligand, a new vaccine can be developed that will be effective against a greater number of strains and in all patient groups.

Sonya Elango – Analysis of low complexity sequence regions in proteins.

Supervisor: Brian Golding
Low-complexity regions (LCRs) are common features of the eukaryotic genome that are characterized by a compositional bias towards a small number of different amino acids. While these regions typically form an unstructured protein product, they have been implicated in a number of functional roles ranging from mRNA stability to circadian rhythm duration. However, despite their ubiquity and roles in a variety of diverse areas, LCRs are very poorly characterized. While this is the case, LCRs share certain characteristics with microsatellite repeats, a relatively well-characterized class of sequences. The similarities between the two types of sequences can therefore provide some insight into the characterization of LCRs. In particular, a number of different models of microsatellite evolution have been proposed and can potentially be applied to the study of LCRs. This project seeks to provide some insight into LCR evolution and dynamics by performing computer simulations that apply these models to LCRs. In order to do this, the Approximate Bayesian Computation (ABC), a statistical technique, will be used to determine which models of microsatellite evolution are relevant to LCRs. Two parameters that the ABC manipulates in this study are the rates of replication slippage and point mutation. The performance of these forward simulations in the C programming language will yield certain values for these parameters in each model. Certain models can therefore be eliminated based on the biological feasibility of these parameter values. Based on previous experimental results, we predict that certain models that take into account factors such as length dependency and rate asymmetry will be better at describing LCR behaviour. These models include Kruglyak’s proportional slippage model and Walsh’s linear birth death chain. In applying these models to LCRs, we can gain some insight into the characterization of these regions. As LCRs are so ubiquitous and functionally diverse, learning more about their evolution, dynamics, and life-cycle may be very important.

Rodrigo Narro-Perez – Investigating Groundwater-Surface water interactions in Silver Creek, Georgetown, Ontario.

Supervisor: Carolyn Eyles
High altitude animals face a variety of environmental stresses associated with their habitat. One of the most essential hurdles organisms face when living in these environments is developing methods for improving oxygen transport and maintaining aerobic capacity throughout the body. This study assessed the effects of both phenotypic plasticity through acclimation, and genetic adaptations by living at high altitudes over extended periods of time. Our experiment used two populations of lab-reared deer mice (lowland Lincoln mice and highland Mount Evans mice), and acclimated both either to sea level normoxic conditions or high-altitude hypoxic conditions (4300m) for 6 weeks. Histological sections of the gastrocnemius muscle, the heart muscle, and the adrenal gland were used to examine various differences that may have been present either between the populations, or arisen through acclimation. Highland mice have shown to have greater capillary density, capillary surface area, and capillary to fibre ratio compared to their lowland counterparts. Highland mice also had a greater aerial and numerical density for Type I and Type IIa fibres, whereas lowland mice had greater Type IIb aerial and numerical density.  Data for the heart sections and adrenal sections is forthcoming. These preliminary results already begin to show some of the numerous genetic adaptations acquired by the Mount Evans mice through living at high altitudes.

Thilakshan Arulnesan – A Computational Study of Superbubbles.

Supervisor: James Wadsley
Computational astrophysics explores the boundary between theory and experimentation. Unlike many scientific disciplines, the processes that are studied by astrophysicists are difficult to observe experimentally. The majority of evidence in the field is from electromagnetic radiation (of various wavelengths), which is observed primarily through a wide variety of telescopes. Computer simulations provide an intermediary type of experimentation, that is they can be used to explore the limits of a theory directly by allowing computational physicists to create their own unique environment in which they can test the repercussions of a certain theory. I will discuss how computational techniques to understand a very specific problem in astrophysics pertaining to the study of superbubbles.

Super bubbles are large (aprox. 30 pc or 100 ly radius), relatively spherical shells of gas that can be observed optical or ultraviolet telescopes. Their structure was first described in two foundational papers by Weaver et al. in 1975. There are several questions regarding the feedback mechanisms involved in the creation of these bubbles. Specifically, a larger intensity of x-rays are produced by these bubbles than predicted by the theory. This questions the understanding of how energy is transported in the super bubble.  One specific portion that is not well understood is the amount of energy transport through conduction of heat and the transport of matter when in the super bubble phase.  The hypothesis that will be examined is that a pressure driven gradient is produced that is in the opposite direction of the temperature gradient which accounts for the excess mass found in the super bubbles.

The hypothesis has been examined in the past using a large scale hydrodynamical simulation which was not originally intended for this purpose (these simulations run very large scales spanning hundreds of thousands of galaxies, resulting in low resolution for this type of problem). In order to verify the results, a mesh-refinement driven code written in C++ called ENZO was used. A very specific one dimensional profile of the super bubble was used in order to understand the conduction process and mass transfer.

Gillian Criminisi – Perceptions of the group selection process in group work.

Supervisor: Sarah Symons
The ability to work in collaboration with others is a highly valuable skill in the workforce. Giving students the opportunity to complete group projects can not only foster these valuable skills, but has also been shown to lead to higher-quality work and a greater understanding of course content. However, group projects can be detrimental if groups are dysfunctional. It is important to consider how these groups are composed in order to increase productivity, decrease conflict, and maximize the potential for learning. Although some studies have compared the effectiveness of certain group formation methods in terms of students’ grades, there has been little research concerning the actual preferences of students. Furthermore, most studies have compared only the random assignment (students are randomly assigned to a group by the instructor) and self-selection (students choose their own groups) methods of group formation. In this study, 45 students in the Integrated Science program at McMaster University completed an online questionnaire which examined four different methods of group selection (random assignment, self-selection, selection based on interest in a common topic, and selection based on answers to a survey), and asked students which methods they prefer, why they prefer them, and whether their preferences have changed over time and with experience. Analysis of the results determined that many students prefer choosing their own groups because they believe that this results in the least amount of conflict, the best communication and the most trustworthy group members. Students also favour group selection based on interest in a common topic as they believe this method is the most fair, the most accurate reflection of how teams will be formed in the workforce, and the best method to use when students are unfamiliar with each other. Moreover, it was found that although students preferred common interest in a topic method in their first year of study, they came to favour the self-selection method as they completed the program. Finally, differences were found between students’ preferences based on their gender and their current year of study. The results of this study are relevant to any educational institute that utilizes group work, but especially small programs or classes in which students will become familiar with each other over time. Further studies should examine whether the same pattern of preferences is found in larger programs and in other faculties.

Ho Jun Eom – Defining an absolute reference frame for “clumped -isotope” geochemistry at McMaster University.

Supervisor: Sang-Tae Kim
Stable isotope thermometry extrapolates the past temperature of formation of various materials such as corals, ocean sediments, and speleothems, by examining the stable isotope signature. Commonly, only the isotopologues with a single isotope are examined due to their relative abundance. Species containing more than one rare isotopes exist in the order of 10 parts per million and were ignored or treated with unverified assumptions in the past. However, these ‘clumped isotope’ have become increasingly important because they provide temperature information of the past without the knowledge about the initial isotopic composition of the formation environment unlike traditional carbonate-water oxygen isotope exchange thermometry.

As a result of increased measurements involving multiply substituted isotopologues, especially the abundance of mass-47 CO2 isotopologues (, an establishment of an absolute reference frame for the dual-inlet stable isotope mass spectrometer at McMaster Research Group for Stable Isotopologues (MRSI) would benefit future studies by producing results which are comparable to results from other laboratories and which empirically accounts for analytical artifacts.

An empirical transfer function is a mathematical relationship between the input of the stable isotope measurements in relative terms to a standard sample and the output value which is in relation to an absolute reference frame. We define the absolute reference frame using thermodynamically equilibrated samples of CO­2. The samples are equilibrated through isotope exchange with water at constant temperature. ∆47 signal depends only on the temperature at which equilibration is achieved because it identifies the clumped isotopes of CO2. Therefore, given the equilibration temperature, the theoretical ∆47 value is known. The linear relationship between the relative ∆47 compared to the working reference gas with the known theoretical values defines the empirical transfer function.

The field of clumped isotope geochemistry is a nascent field with many potential applications. It has provided more temperature information in environments where such information was unavailable previously. The oxygen isotope thermometry is still widely used to this day due to its simplicity; clumped isotope thermometry is a viable technique to supplement in problems where there is not enough information about the past environment to apply the pre-existing method. The establishment of standardized method for clumped isotope data reporting, which uses the common absolute reference frame and accounts for various artifacts of measurement, will improve the quality of data available for examination.

Meiko Peng – Exploring the Bio-activity of Small Molecules Targeting Gram-Negative Cell Walls.

Supervisor: Eric Brown
The advancement of public health is continuously being challenged by the rise and spread of antibiotic resistant pathogens. Historically, antibiotics have been game-changing for modern medicine; however, the pace of antibiotic discovery has slowed significantly in recent years. Consequently, new strategies for developing antibiotic screens and connecting lead compounds to their mechanism of action are necessary to revitalize this field. In recent years, infections by drug-resistant Gram-negative bacteria have become more problematic. Lower intrinsic permeability to antibiotics and abundance of multidrug efflux pumps make these bacteria challenging to combat and thus highlight the pressing need to focus on targets in Gram-negative microorganisms.

We report here the first use of high-throughput bacterial microscopy coupled to multicopy suppression as a forward genetic method to determine cellular targets within the Gram-negative bacterial cell wall. A screen of 140,000 small molecules for growth inhibition of Escherichia coli led to the identification of 1197 actives, which were assessed via high-throughput microscopy to identify compounds that altered bacterial morphology. This secondary microscopy screen involved batch processing each micrograph through the use of customized image analysis software that can identify and quantify morphological parameters of each individual cell for identification of significant hits.

Using this approach, we discovered 20 small molecules that caused morphology changes to the normally rod-shaped E. coli, of which 5 were shortlisted for further investigation. These 5 small molecules were systematically screened against the KEIO deletion and ASKA overexpression collection for sensitivity and suppression, respectively. This approach led to the discovery of C3, a molecule structurally similar to furazolidone, but may indirectly interfere with bacterial cell division by targeting threonyl-tRNA synthetase (thrS). This method of screening not only represents a novel strategy in the identification of drug leads but is also an exciting new way to screen for cell wall drugs. Given the global emergence of antibiotic resistant bacteria, the implication of both new screening techniques, and end results with new drug candidates, is a tremendous step forward in drug discovery.



Hannah Dies – Amyloid-beta protein interactions in biological membranes.

Supervisor: Maikel Rheinstädter
The symptoms of Alzheimer’s disease do not manifest in most patients until after the age of 60, suggesting a strong link between age-related changes in the composition of brain tissue and progression of the disease. In particular, age-related changes in cholesterol and melatonin levels are currently highly discussed in the literature. We studied the interaction between two amyloid-b peptides, amyloid-b1-42 and amyloid-b25-35, with anionic lipid model membranes containing cholesterol and melatonin. Our high resolution X-ray setup enabled us to determine the molecular structure of the membranes with sub-nanometer resolution in-situ, under physiological conditions. We determined the location and orientation of the molecules in the bilayers, and their lateral organization in the plane of the membranes. Both peptides were found to embed in the membrane core, which is believed to be crucial for the formation of oligomers. Moderate levels of cholesterol (30mol%) led to the formation of cholesterol plaques in the anionic membranes. The Ab25-35 peptides were found to strongly interact with the membrane, displacing cholesterol molecules from the lipid regions into the plaques and increasing the total fraction of plaques in the membrane. The melatonin molecules were found to reside in the head group region of the membranes and increase the fluidity of the anionic membranes, in effect inhibiting the insertion of Ab25-35 into the hydrocarbon core of the bilayers. We, therefore, present direct experimental evidence for an interaction between Ab peptides, melatonin and cholesterol on the level of the cell membrane and suggest an important role for age-related effects in membrane-peptide interactions.

Alexander Young – Hypoxic Tolerance in Fish.

Supervisor: Graham Scott
Human activity has the potential to drastically alter species diversity, distribution, and abundance in a local environment. This activity can have similar effects on a global scale, exposing organisms to a wide range of novel environmental stressors. Currently increases in both mean temperatures and the frequency of extreme weather events have been noted worldwide due to anthropogenic activity, and current predictions suggest that this trend will continue. In light of this, the development of reliable models that can accurately predict the effects of anthropogenic environmental changes on organisms is paramount in order to determine which species will be most severely affected. These models must be holistic, as environmental stressors are not isolated, and instead interact with one another resulting in additive and multiplicative effects. The temperature and oxygen content of aquatic environments provides an excellent example of the synergistic nature of stressors, as both have profound effects on the biological activity of resident organisms. In particular, increases in water temperatures have been shown to raise the metabolic rate of aquatic organisms, and hypoxic conditions have been linked to the mass die-off of fish around the globe. These events are likely to co-occur due to eutrophication, which is stimulated by high temperatures, and leads to an increase in algal abundance, exacerbating hypoxic conditions. The purpose of this study is to better understand the relationship between hypoxia and temperature and their role as interacting stressors in aquatic environments, particularly in closely related species with overlapping habitats. Two species in the family Centrarchidae, specifically bluegill (Lepomis macrochirus) and pumpkinseed (Lepomis gibbosus) sunfish, and an infertile hybrid of the two, were examined. These three groups of fish have been shown to inhabitant a similar geographical range; however, at northern latitudes a higher population density of pumpkinseed sunfish has been noted (Farwell et al., 2006). This study aims to provide evidence regarding the species distribution of these species and if increased tolerance to hypoxia explains the increased northern environmental by measuring oxygen consumption, a proxy for metabolic rate, using respirometry trials at 15 and 25ºC. These results will also highlight the effects of hypoxia and temperature as interacting stressors in closely related species, aiding the creation of more reliable and holistic climate change models.

Paras Patel – Hypoxic adaptations of deer mice.

Supervisor: Graham Scott
Animals living at high-altitudes face a variety of environmental stresses associated with their habitat. One of the most essential hurdles organisms face when living in these environments is developing methods for improving oxygen transport and maintaining aerobic capacity throughout the body (Storz et al., 2010). This study assessed the effects that phenotypic plasticity and genetic adaptation have on the physiology of animals living at high-altitudes. Our experiment used two populations of lab-reared deer mice (lowland Lincoln mice and highland Mount Evans mice), and acclimated both to either sea level normoxic conditions or high-altitude hypoxic conditions (4300m) for six weeks. Histological sections of the gastrocnemius muscle and the heart were used to examine various differences that may have been present between the populations, or arisen through acclimation. Highland mice have been found to have greater capillary density, surface area, and capillary-to-fibre ratio compared to their lowland counterparts. Highland mice also had a greater aerial and numerical density for Type I fibres, whereas lowland mice had greater Type IIb aerial and numerical density.  Heart sections revealed that lowland mice increase their capillary density and capillary surface area after acclimation, whereas highland mice show no change. These results confirm that the highland Mount Evans mice have developed numerous genetic adaptations to succeed in life at high-altitudes.

Kristin Daoust – Exploration of the effects of environmental variability on diversity and stability of ecological communities using a long-term data set on a natural model meta-ecosystem.

Supervisor: Jurek Kolasa
Most experimental and observational studies in ecology are conducted at relatively modest spatial and temporal scales, yet ecologists are becoming increasingly aware that the structure and dynamics of local populations and communities may reflect processes occurring at larger spatial scales over longer periods of time. Despite understanding the significance of spatial patterns for species interactions, there has been very little attention given to the combined effects of spatial patterns and temporal variability, also known as spatiotemporal relationships. A pair of indices, Predictability (PST) and Temporality (T), created by Matthew Hammond from McMaster University can help to reveal the extent of the spatiotemporal environment affects biotic dynamics. PST and T have a generic structure and can be applied to combinations of factors other than just space and time. In the past, these indices have not been used to examine other factors; however, this study investigates not only the relationship between space and time, but also the relationship between the species of interest and time (PCT), as well as species and space (PCS). In hope to find more details about these relationships, data from 49 rock pools in Jamaica near Discovery Bay was used to calculate the three index pairs. For, space by time, a principle component analysis and a backwards multiple regression was used to determine that PSTand niche breadth had a significant, positive relationship, meaning species that were generalists were more easily predictable than specialists. For, species by time, a general linear model (GLM) was used to determine the interactive effect of species richness and the variability of the habitat (HVI) had on PST, which showed that richness was poor as a predictor for the dynamics of the ecosystem when HVI was high. These results have global implications for predicting the behaviour of ecosystems at a larger scale. Future studies should focus on recreating these results with a larger population of a species important to humans in order to better understand the effects climate change could have on species in certain ecosystems.

Sheridan Baker – Effects of Cu and hypoxia on fish.

Supervisor: Grant McClelland
Anthropogenic influences are presently causing severe ecological problems. Not only is preserving ecological diversity important; but these problems can also have negative impacts on our society as many ecosystems are of economic and commercial value. To protect fragile ecosystems from these negative anthropogenic effects, it is important to understand how organisms react and respond to changes in their environment. One area that requires investigation is aquatic metal toxicology. Industrial activity such as mining can lead to leaching of metals such as copper into aquatic environments. Additionally, eutrophication, which can lead to aquatic hypoxia, is another ecological issue that has increased in prevalence in recent years. This project examined how chronic copper exposure affects the acute ventilatory drive of killifish (Heteroclitus fundulus). In this study, killifish were exposed to 100 μg Cu/L for 96 hours and then exposed to normoxia, hypoxia (20% O2 saturation), or 100 μM ammonia. Their respiratory frequencies were measured using a glass tube with electrodes at each end that measured changes in impedance associated with gill movements of the fish. It was hypothesized that chronic copper exposure would cause the fish to be deficient in their respiratory response to these stimuli when compared with control fish unexposed to copper. Preliminary results have agreed with this hypothesis. Additionally, hypoxia caused a larger increase in breathing rate than ammonia exposure. This indicates that both aquatic copper pollution and hypoxia can pose ecological problems to aquatic environments and precautions should be taken in the future to avoid pollution of valuable aquatic ecosystems.

Charles Yin – Signal Transduction Mechanisms of the Macrophage Scavenger Receptor A (SR-A).

Supervisor: Dawn Bowdish
The macrophage scavenger receptor A (SR-A) is a phagocytic receptor that plays a key role in host defence and the regulation of tissue homeostasis. It is capable of recognizing a diverse range of endogenous and exogenous polyanionic ligands and mediates the initiation of ligand-specific responses. To date, there is little known about the signal transduction pathways employed by SR-A or even which signaling intermediates it recruits upon ligation. A recent study by Yu et al. (2011) demonstrated that SR-A interacts directly with TRAF6 upon stimulation with LPS – the first study to identify a signalling adaptor for SR-A. However, it remains unclear how SR-A is able to interact with TRAF6 since the only putative TRAF6-binding motif on SR-A is located on the extracellular portion of the receptor. The aim of this project is to determine the means through which SR-A is able to recruit TRAF6. To this end, we generated a SR-A mutant lacking its TRAF6-binding motif. Although an uptake assay using fluorescent beads showed that this mutant showed equal capacity to bind and internalize ligand, it proved unable to down-regulate TLR2-driven pro-inflammatory signaling to the same extent as the wild type receptor when tested using an NFκB reporter assay. Therefore, SR-A must somehow make its TRAF6-binding motif accessible to the cell cytoplasm. We propose that SR-A undergoes an inversion of its orientation across the membrane bilayer upon stimulation with a ligand, causing the previously extracellular portion of the receptor to be now accessible to proteins located in the cell cytoplasm, such as TRAF6. Should this model be accurate, then the molecular biology of the cell membrane would prove to be much more complex than what we have previously imagined.

James Seong Jun Han – Pattern associated modelling for discovery of novel protein motifs in the macrophage receptors.

Supervisor: Dawn Bowdish
Class A scavenger receptors, including scavenger receptor A and macrophage receptor with collagenous structure, are surface proteins that bind modified endogenous ligands and bacterial components. As these receptors play a key role in innate immunity, they are associated with a range of infectious diseases such as pneumonia. Thus, understanding the physical properties of these receptors will provide insight for the development of novel therapeutics. Macrophage receptor with collagenous structure is expressed to mediate pathogen recognition and is amongst the least well-characterized members of the scavenger receptor family. Our investigation addresses the development of an efficient and highly accurate bioinformatics technique through combinatorial usage of Aligned Pattern Clustering, and Multiple Sequence Alignment Pattern Retrieval Program for novel motif discovery in the macrophage scavenger receptor. By utilizing scavenger receptor A and its validated motifs as a model, we successfully validated the feasibility of this bioinformatics technique with a minimum sensitivity of 54% and positive predictive value of 60%. As a subsequent validation of the newly discovered motifs, the top sub-sequences of macrophage receptor with collagenous structure, selected by both programs, were analyzed for their biological functionality. Here, we propose that RGRAE, VFCRMLG, EDAGVE and WGTICDD motifs in the scavenger receptor cysteine-rich domain play an important role in pathogen recognition.

John Rawlins – How are screening policies made?

Supervisor: Julia Abelson
Prostate cancer screening using the PSA test is one of the most controversial issues regarding health policy in Ontario today. While the vast majority of large scale, randomized control trials dealing with this subject matter continue to produce results which show that screening for prostate cancer with the PSA test causes no significant reduction in prostate cancer specific mortality rates and, as such, Ontario continues to refuse any sort of PSA screening coverage, men across the province continue to seek out, and undergo PSA based cancer screening. Several factors may be influencing this scientifically unfounded desire, including social media claims, religious beliefs, preconceived notions of the benefits of screening, the view of clinical practitioners, and, perhaps most importantly, the stance taken by traditional media outlets (the outlet by which people most commonly hear about health policy issues). In order to assess the potential impact that Ontario’s traditional media sources may be having on the public’s opinion of, and desire for, PSA testing, a media content analysis was completed. All articles on the subject matter of PSA testing published in the National Post, Toronto Star, and Globe and Mail between January 1, 2006, and February 14, 2014 were analyzed in order to determine the stance taken by these sources on PSA testing, and how it may be impacting public desire for screening. The results of this study show that the majority of articles published frame PSA testing in a positive manner. Furthermore, the types of articles published (news vs editorial), the articles specific time frame of publication, and the variance in persuasiveness between the positively and negatively framed articles may also contribute toward the disconnect between the scientific consensus and public desire for PSA testing.


Amna Hyder – Random Networks.

Supervisor: Maung Min-oo
Pattern recognition in machine learning aims to reduce a large dimensional data set to a lower dimensional manifold contained within it in order to eliminate redundant information. For example, a photograph with $n^2$ pixels, corresponding to an n by n image, would have a dimensionality given by $\R^n^2$, however, this is likely higher than the intrinsic dimensionality of the image, which is dependent on the degrees of freedom of the camera. Principle Component Analysis is one of the most popular methods used to reduce the dimensionality of data sets by producing a set of linearly uncorrelated variables in a new coordinate system. However, PCA does not consider the intrinsic structure of the manifold, which motivates us to use another method from graph theory that would take such topology into account. In graph theory any abstract object or collection of objects can be represented by vertices, and the links between them by edges. The combinatorial laplacian is a matrix representation of the graph that preserves local information and considers the geometry of the manifold contained within it. The first three eigenfunctions of the combinatorial laplacian for two different data sets, protein-protein interactions and carcinogenicity of hydrocarbons are plotted to determine key features. Roughly twenty thousand different properties of 303 different polyaromatic hydrocarbons are plotted along with carcinogenicity to determine which features play the biggest roles in the toxicity of these compounds. The identification of patterns in this data set through the eigenfunctions of the combinatorial laplacian may help determine which features can be used as predictors for polyaromatic hydrocarbon carcinogenicity.

Eric Hempel – Investigation of geometric models in general relativity.

Supervisor: Maung Min-oo
In 2006, the community of Georgetown located in the town of Halton Hills, Ontario was identified as one of Ontario’s ‘Places to Grow’. A population increase of 50% is expected by 2031.  Georgetown extracts their water supply from various well fields that are supplied with water from aquifers hosted in Quaternary sediments infilling a buried bedrock valley. The Cedarvale Well Field aquifer is located beneath the modern fluvial systems of Silver Creek and Black Creek. There is a significant amount of water that is lost from Silver Creek into the Cedarvale Well Field, although little is known about specific location and or magnitude of this groundwater-surface water interaction (SW-GW). An understanding of this interaction is of importance with the prospect of increasing pumping rates that will be incurred by an increase in fresh water demand that is expected with the community’s growth. A framework was constructed using sedimentological and thermal data collected in Silver Creek. Six dominant facies types (Facies 1: pebble/granule dominated facies; Facies 2: gravel dominated facies with a venner of fine-grained sediments less than 1 cm thick; Facies 3: any gravel dominated facies  with a venner of fine-grained sediment between 1cm and 10cm thick; Facies 4: any gravel or sand facies covered with organic material; Facies 5 is a sand facies; Facies 6: organic material deposited in areas of minimal flow) were characterized and illustrated into a streambed sediment map for an extend of 600m. FA Temperature Stick which was constructed specifically for this project was used to make a detailed thermal survey which measured the streambed temperatures at a depth of 10 cm. Halton Region provided the thermal data from groundwater monitoring wells for the years 2011-2013. Analysis of the groundwater temperatures consisted of the use of Microsoft Excel where trends were observed for the two-year periods. Analysis of the thermal data also included simple calculations and transformations of the plotted data. A thermal map was made from the thermal surveys which illustrates the spatial variability of the streambed within a small section of the stream. From these results it was apparent that heat is moving at varying rates throughout the subsurface as well as near the streambed. Various areas were identified as key areas of GW-SW interactions. To further expand this study it is recommended that various thermal surveys be made throughout the Creek during different times of the year to provide both duplicability and reliability of the data.

Madison Reid – Student Satisfaction with Blended Learning Modules Promoting Deep Learning.

Supervisor: John Maclachlan
Online and blended learning initiatives are becoming increasingly common in post secondary institutions. Despite their increased use, the effectiveness of online and blended learning techniques is not well understood. This project investigates online learning modules recently implemented in the School of Geography and Earth Sciences at McMaster University, to be used in conjunction with traditional lecture material. The goal of these modules was to promote deep learning in students.

The main objective of this project was to assess student satisfaction with the blended teaching approach used. Data collection used a survey designed to gage student satisfaction with the modules that composed the online portion of the blended approach. The survey was also designed to investigate the effectiveness of the modules in promoting deep learning approaches. The survey was introduced to each class that used the online modules, and participation was optional. There were 64 responses. Student satisfaction with the modules varied from positive to negative extremes, however most responses were positive. To further understand students’ opinions, responses were analyzed by faculty, year, and other categorical factors. In addition, indicators of deep learning such as comprehension of material and interaction with peers and faculty were discussed.

The results of this study will serve as an example illustrating some possible methods for reporting the effectiveness of blended and online teaching techniques. While the data from this study are case specific, online and blended learning is becoming increasingly prevalent in educational institutions globally. As such, insights of methodology and analysis could be applicable on a much greater scale. In addition, the conclusions of this study could help support the implementation of these modules at McMaster University by identifying existing strengths and weaknesses, and suggesting improvements.

Sarah Drohan – Childhood Disease Dynamics in NYC.

Supervisor: David Earn

The spread of infections between isolated populations plays an important role in determining overall temporal disease dynamics. One interesting phenomenon displayed by spatial disease dynamics in a metapopulation is that of synchrony. Defined as “the coincident change in incidence or abundance between populations” , the level of synchrony can reveal pathways for disease transmission. In a sense, synchrony provides information about whether populations undergo epidemics at the same time.

In 1999, Rohani et al. studied the synchrony of whooping cough and measles epidemics between the 50 largest cities in England and Wales using historical data and mathematical models. For whooping cough they found that epidemics synchronized after vaccination while for measles they found that epidemics desynchronized after vaccination. Our research aims to further explore changes in synchrony due to vaccination using both statistical and modelling approaches on a smaller spatial scale. This goal will be accomplished by approaching the problem from two different perspectives.

First we conduct a statistical analysis of historical disease data. Case reports for six childhood respiratory infections (chickenpox, rubella, measles, mumps, scarlet fever, and whooping cough) for the 30 health districts in New York City have been acquired. This data set is particularly beneficial since the data spans from 1958 to 1976 and vaccination regimes for rubella, measles, and mumps were implemented in 1969, 1963, and 1967 respectively. We use a variety of statistical tools such as Mantel correlograms and the Pearson correlation coefficient to estimate the change in synchrony for all six childhood respiratory infections in NYC. Preliminary results indicate that synchrony decreased for all diseases over the 20-year time period, however no striking shift after vaccination was observed.
The second perspective uses a mathematical model to investigate the effect of vaccination on synchrony. The most common model used to study infectious diseases is the SIR model, which stands for susceptible-infected-recovered model. This defines a system of differential equations governing the movement of individuals between compartments. Using a modified Gillespie algorithm to simulate the stochastic system we will attempt to determine the effect of vaccination and epidemiological coupling on synchrony.

Not only does the relationship between synchrony, spatial spread, and vaccination provide interesting results about dynamical systems, but the outcomes can also be generalized to aid in developing control strategies and vaccination regimes for infectious disease epidemics.

Piotr Roztocki – Speciation and sorption of Np(IV) and Np(V) in brine.

Supervisor: Shinya Nagasaki
The use of nuclear reactors to generate electric power leads to the creation of intermediate and high level long-lived wastes, which must be stored or disposed of in a manner that prevents radionucleotide migration prior to the material’s radioactive decay. Finding safe methods of disposal is a challenge to all countries using nuclear energy, among which is Canada. One of the proposed methods of disposal is the use of deep geological repositories, which could store the spent fuel until it decays. Neptunium (Np) is one of the key nucleotides to dominate nuclear fuel disposal safety assessments due to its long half-life (2.14 million years) and high mobility in shallow groundwaters, which make it a major contributor to total emitted radiation. The speciation of Np and other actinides present in waste has been largely studied in fresh and saline water. However, groundwaters in Canada and especially in southern Ontario are brine with an ionic strength ten times higher than of seawater. The speciation of Np in brine is not clear at the present and there is a possibility that Np complexes not formed in fresh and saline water could form in brine. With multiple repository sites already in Canada and southern Ontario, and potential future ones being developed in regions with brine groundwaters, understanding the speciation and sorption of Np is key to a conclusive safety assessment. An important factor considered in such safety assessments is sorption, the extent to which radionucleotides attach to mineral surfaces, important in assessing whether an actinide species would be highly mobile in groundwater and eventually enter the biosphere, or be filtered out by subsurface minerals. The sorption in turn is significantly influenced by actinide speciation, namely, the chemical species an actinide is most likely to be present as. Our study aims to use surveys of current literature / chemical thermodynamics databases and geochemical modeling codes to extrapolate Np speciation and sorption from fresh and seawater to brine conditions. PHREEQC, a platform for aqueous geochemical calculations, was used for aqueous modelling of Np speciation in Canadian groundwaters. As this type of actinide chemistry research is still in its infancy, sensitivity and predominance analysis results point to the most important modelling parameters to be investigated further by engineers and scientists alike. This type of research is an important step towards a complete safety assessment of nuclear waste management procedures in the context of Canadian geochemistry.

Kaian Unwalla – What Was I Just Thinking About? The impact of forgetting negative words on mind wandering tendency.

Supervisor: Scott Watter
Previous research in the Cognitive Science Laboratory has shown that the temporal focus during mind wandering episodes varies depending on the emotional quality of the stimuli presented during the task. Specifically, research has shown that presenting participants with negative words induces a retrospective focus during mind wandering episodes, without increasing the frequency in which mind wandering episodes occur. This effect is strongest in those who score high on measures of negative mood. Such a finding is not surprising, considering the consistent finding that depressed individuals are biased toward the processing of negative stimuli relative to normal controls. Interestingly, work within the forgetting literature has shown that asking participants to substitute negative associations with neutral associations can act to buffer individuals from the potentially damaging effects of a negative environment.
Combining these two lines of research, the purpose of the current study was to investigate whether performing a substitution procedure reduces the influence of negative stimuli on the temporal focus during mind wandering episodes, especially in those who report a high level of negative mood. Prior to the experiment, participants completed the Positive and Negative Affect Schedule, the Beck Depression Inventory, and the State Trait Anxiety Inventory, to assess their level of negative mood and anxiety. Subsequently, they completed the Think/No-Think task to train the forgetting of negative associations. Finally, participants completed the Sustained Attention to Response Task (SART) to measure their level of mind wandering, using the negative associations as task stimuli. Thought probes were inserted pseudo-randomly throughout task performance to determine whether participants were mind wandering, and if they were, the focus of their off-task thoughts. We predict that those who score higher in negative affect will be more likely to engage in retrospective thought when perceiving negative words, relative to those with a low score, but that this effect will diminish after completing a substitution task of the negative words. This finding would provide evidence for a method to counteract rumination in those with a negative mood.

David McDonough – Effects of online Discussion Participation on Course Performance.

Supervisor: Geoff Norman & Kelly Dore
The advent of the Internet in the 1980s prompted many researchers and educators to see the huge benefit widespread interconnectivity could have to education. Since that time, leaps and bounds have been taken in the realm of online education, from content management systems supplementing university and college courses (i.e. Avenue to Learn at McMaster University) to massive open online courses (MOOCs) free to anyone with an Internet connection (i.e.

While debate continues between whether online education or traditional instruction is better at promoting and facilitating learning, it cannot be denied that there are inherent differences between the two. One important aspect of online education not present in the traditional classroom is the discussion forum, or more specifically, the asynchronous discussion forum. While a dialogue can be established in a classroom with ease, there is no transcript readily available for review afterwards, whereas with the asynchronous online discussion forum, each student’s contributions are recorded and available for re-examination. This transcript of student interaction provides a snapshot of the students’ current understanding of course material – information that is open for evaluation.

The present study evaluates the efficacy of asynchronous online discussion as an assessment tool for student performance. Master of Science in Health Science Education students enrolled in HSEDUC 701: Learning and Curriculum had three different week’s discussions evaluated by four raters on a 6-point scale from B- to A+: one discussion chosen by random from Weeks 1-4, one from Weeks 5-8, and one from Weeks 9-12. Rater’s scores for each student were compared to one another to determine interrater reliability, and compared to multiple-choice test scores to determine validity. It is predicted that reliability and validity coefficients will be promising (>0.70), meaning evaluations of asynchronous online discussion are an accurate representation of student course performance.

John Taehun Kim – The Role of UPR in Modulation of Inflammation in COPD. 

Supervisor: Kjetil Ask
Chronic obstructive pulmonary disease (COPD) continues to be a leading cause of death. The major causes of COPD are cigarette smoke, pollutants, and occupational exposure to dust. Of all lung diseases, COPD is believed to be the most associated with air pollution exposure. Alveolar epithelial cells, and subsequently the NLRP3 inflammasome, are activated in response to air pollutants. As a result, they have been implicated in COPD suggesting that they have a pathogenic role in disease. Recent research has highlighted a role for the unfolded protein response (UPR) in activating the inflammasome in the progression of COPD. Furthermore, the activation of alveolar epithelial cells and the inflammasome in response to UPR provides the ideal system in which to study COPD. The exact mechanisms, by which the exposure to environmental agents leads to COPD, have not been well elucidated. Lipopolysaccharide (LPS) is an endotoxin on the surface of gram negative bacteria that is commonly found in air pollution and organic dust. LPS can activate the inflammasome through a signalling cascade. The current study examines the effects of UPR activation on LPS-indcued inflammation. Human alveolar epithelial cells (A549) were exposed to 800 ng/mL LPS for 2 hours to induce inflammation. A549 were also treated with UPR inhibitors, STF-083010, tauroursodeoxycholic acid (TUDCA), and 4-(2-aminoethyl)-benzenesulfonyl fluoride (AEBSF). Subsequently, cell viability and levels of cytokine production were measured. We hypothesize that the inhibition of UPR will prevent inflammation induced by LPS exposure. The present study could provide alternative inflammatory mechanisms in an alveolar epithelial disease model.