Welcome to the upgraded MacSphere! We're putting the finishing touches on it; if you notice anything amiss, email macsphere@mcmaster.ca

About MacSphere

MacSphere is McMaster University's Institutional Repository (IR). The purpose of an IR is to bring together all of a University's research under one umbrella, with an aim to preserve and provide access to that research. The research and scholarly output included in MacSphere has been selected and deposited by the individual university departments and centres on campus.

To contribute to McMaster's Institutional Repository, please sign on to MacSphere with your MAC ID.

If you have any questions, please contact the MacSphere Support Team.

Students wishing to deposit their PhD or Masters thesis, please follow the instructions outlined by the School of Graduate Studies.

Discover

Communities in MacSphere

Select a community to browse its collections.

Recent Submissions

  • Thumbnail Image
    Item type: Item ,
    Evaluating Biomolecular and Biochemical Characteristics of Anti-Platelet Factor 4/Heparin Antibodies in Serological Subtypes of Heparin-Induced Thrombocytopenia
    (2026) Kwok, Sarah
    Heparin induced thrombocytopenia (HIT) is an adverse reaction to heparin (a commonly used anticoagulant), characterized by a low platelet count and potentially catastrophic thrombosis. Resembling HIT is vaccine-induced immune thrombocytopenia and thrombosis (VITT), which arose in response to the COVID-19 pandemic due to SARS-CoV-2 adenoviral vector-based vaccines. Both of these immune-mediated disorders are caused by antibodies against the chemokine, platelet-factor-4 (PF4/CXCL4), which activates platelets through the FcγIIa receptor, leading to the release of procoagulant platelet microparticles, reduction in platelet count level, and a high risk of venous and arterial thrombosis that can be severe and even fatal. There are two serological subtypes of HIT based on their platelet activation patterns in functional assays: classical heparin-dependent HIT (HD-HIT) antibodies that require heparin for platelet activation, and autoimmune heparin-independent HIT (HI-HIT) antibodies that do not require heparin for platelet activation. HI-HIT is a more severe subtype associated with higher rates of disseminated intravascular coagulation (DIC), more severe and prolonged thrombocytopenia (platelet count of <20 × 109/L), and persistence of symptoms after heparin cessation. Previous research in our laboratory found that HD-HIT and HI-HIT bound to different regions on PF4. Interestingly, in VITT, the precise anti-PF4 antibody epitopes correlated with clinical outcomes, such as whether the rare event of cerebral venous sinus thrombosis (CVST) occurred. Our research suggests that the precise antibody epitope targeted may be important for determining clinical outcomes in anti-PF4 disorders; by extension, we want to investigate whether this observation exists in HD-HIT and HI-HIT. In this thesis, we conducted an in-depth comparison of HD-HIT (n=16) and HI-HIT (n=20) subtypes at the level of anti-PF4/heparin epitopes, antibody binding affinity, antibody titre, and clinical outcomes. We used established methods to determine the anti-PF4/heparin epitopes for a large cohort of HIT patients to expand upon earlier preliminary work. We found that the combined HIT binding site on PF4 (n=36) was distinct from previously identified epitopes of monoclonal anti-PF4 antibodies, including the KKO, 5B9, VITT, and 1E12, as well as the heparin binding site. By gaining a deeper understanding of anti-PF4/heparin binding sites in HIT, we can potentially develop diagnostics that can target this shared binding region. Additionally, we found that HD-HIT and HI-HIT antibodies bind to very similar regions on PF4, which was contrary to our expectations, but agrees with our recent finding that HIT is driven by a pathogenic monoclonal IgG antibody. Rather than differences in anti-PF4/heparin epitopes, we provide evidence that HI-HIT antibodies have higher antibody affinities and titres compared with HD-HIT. Additionally, we found one amino acid residue (P37) that was significantly different in HIT patients with a platelet nadir above (n=6) and below (n=7) 50 x109 platelets/L, and we found seven residues (D7, L8, T16, V29, A39, K50, I64) that were significantly different between HIT patients with (n=8) and without (n=5) thrombotic complications. Overall, this suggests that separating HIT patients based on their platelet activation patterns in functional assays may not be as important as separating HIT patients based on clinical outcomes.
  • Thumbnail Image
    Item type: Item ,
  • Thumbnail Image
    Item type: Item ,
    Resolving Temporal Graph Functional Dependency Violations
    (2026) Bhandari, Satyagni
    Data is often riddled with errors and noise, from misspellings and missing values, that need to be dealt with to use the data effectively. This kind of preprocessing is very costly; an estimated cost to U.S. businesses of $3 trillion as reported in 2016 by Harvard Business Review. Temporal graphs are used to record changes in graph data, capturing both topological and attribute value constraints and relationships. As data values change, snapshots of temporal graphs may become inconsistent with respect to a set of graph quality constraints that impose topological and attribute value requirements. For example, monitoring patient drug dosages over time involves relationships between attributes such as a patient’s condition, treatment, symptoms and specific dosage values, while adhering to strict dosing requirements over specific time intervals. This thesis studies the problem of resolving violations to a specific class of graph data quality rules called Temporal Graph Functional Dependencies (TGFDs). The thesis introduces a new algorithm called TGFD-Correct, that processes and groups all violations sharing the same antecedent values (with respect to the TGFD), and selects a single (consequent) value for each group that balances two objectives. First, fixing the data (with respect to the TGFD), and second, minimizing changes to the distribution of consequent values from a prespecified objective. As each group is being handled independently, the repair step can be run in parallel across many processor cores, making it fast enough for millions of records. Our method had a 3-point improvement over the best performing baseline that we tested from the literature, whilst maintaining a runtime footprint 17 times smaller. Taken together, the findings show that careful, distribution-aware cleaning of temporal graphs is possible, and demonstrates a path toward more flexible tools that can keep pace with ever-growing, time-stamped data.
  • Thumbnail Image
    Item type: Item ,
    The Numerical Solution of Differential Equations by Third and Fourth Order Runge-Kutta Methods
    (1964-10) Ebos, Frank
    C. Runge originally suggested the numerical methods of solving differential equations which will be examined, and were subsequently improved on by, to mention a few, K.Heun, and W. Kutta. The entirety of these methods have, as a result, been referred to as the Runge-Kutta methods for the numerical solution of differential equations. The first section of the thesis consists of the derivation of third and fourth order Runge-Kutta methods and their respective truncation errors. Notation, definitions, and various concepts are introduced as needed in the various sections. The numerical solutions of differential equations using third order Runge-Kutta methods are then discussed in the second section. Various formulae and relationships are derived here for third order methods. In all numerical tables that follow, the results were obtained using a Bendix Model G-15 Digital Computer. In the third section, one considers fourth order Runge-Kutta methods for the numerical solution of ordinary differential equations. However, in addition to considerations of symmetry, reduction of operations and storage requirements, as examined in section two, one examines a Runge-Kutta method due to Blum which basically modifies a programming procedure. Finally in the last section, one investigates methods due to A. Ralston which minimize a bound on the truncation error derived in the first section. An appendix is also included containing various programs for the Bendix G-15D that have been needed throughout the sections.
  • Thumbnail Image
    Item type: Item ,
    Values and Preferences of Patients with Small Renal Masses: Integrating Evidence and Patient Perspectives
    (2026) Kandi Keleh, Maryam
    Small renal masses present patients and clinicians with preference-sensitive choices among partial nephrectomy, percutaneous thermal ablation, and active surveillance. This thesis set out to generate decision-ready evidence on both short-term burdens and longer-term oncologic outcomes for each option, and to explore how patients value the trade-offs involved. We conducted coordinated prognosis-focused systematic reviews and meta-analyses to estimate strategy-specific probabilities for peri-operative outcomes (major complications, blood loss, hospital stay, procedure time) and for five-year oncologic endpoints (overall mortality, cancer-specific mortality, local recurrence, metastasis). Methods emphasized applicability to decisions: QUIPS for risk of bias; GRADE adapted for prognosis (including a decision-threshold approach to imprecision); ICEMAN to guard against over-interpretation of subgroup signals; and improved conversions of medians/IQRs to means/SDs to avoid distorted weights in single-arm syntheses. We then built and piloted a bilingual decision aid and used structured interviews with a probability trade-off task to elicit values and risk thresholds. Across strategies, short-term burdens differed in ways patients care about, while five-year cancer-specific outcomes were generally favourable; estimates are presented as descriptive prognosis for each option rather than comparative effects, with certainty ratings that make remaining uncertainty explicit. The pilot demonstrated the feasibility and acceptability of eliciting preferences and showed meaningful variation in how patients balance avoiding procedures against small differences in oncologic risk. By integrating rigorous prognosis evidence with empirical preference elicitation, this thesis provides a transparent foundation for shared decision-making and supports conditional, values-sensitive recommendations in guidelines. It also offers practical methods, both statistical and procedural, for future decision-aid development in SRMs and similar preference-sensitive conditions.