Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

While the statistical mechanics of systems in thermal equilibrium is a well established discipline, nonequilibrium systems are fundamentally much less well understood, even though most natural phenomena fall into the latter category. In particular, there is as yet no nonequilibrium analog for the systematic formalism of Gibbs ensembles. Rather than deal with the difficult problem of general nonequilibrium systems, this study is restricted to the steady states of a simple model whose equilibrium properties are well known. The Ising lattice gas displays a number of surprising phenomena when driven into nonequilibrium steady states. This study extends previous work to a more general model with anisotropic interparticle interactions. Using Monte Carlo simulations, we obtain the phase diagram for the model, controlled by the driving field, temperature, and anisotropy parameter α. Under saturation drive, the shift in the transition temperature between ordered and disordered states can be either positive or negative, depending on α ≡ √(𝐽_∥/𝐽_⟂). The possible existence at large α of an additional phase ordered in only one direction is discussed. For finite drives, both first and second order transitions are observed. A novel technique for locating the first order transition line is presented. Some aspects of the phase diagram can be predicted by investigating the two-point correlation function to first order in a high temperature series expansion. However, the series expansion fails to predict even qualitatively the α-dependence of the critical temperature. / Master of Science

No description available.

Indiana University-Purdue University Indianapolis (IUPUI) / The dissertation consists of three essays attempting to assess the impact of expanded health insurance policy on cancer care continuum and measure the unbiased program effects after taking care of mis-measured cancer screening self-reports. The first essay examines the impact of the Affordable Care Act’s Medicaid expansion on time to oral cancer treatment initiation since diagnosis, quality of hospital care such as length of stay in the hospital, planned and unplanned readmissions post-surgery, and care outcome such as ninety-day mortality since surgery. The study uses two-way fixed effects linear model analysis under a difference-in-difference estimation setting to show that Medicaid expansion eligibility reduced overall oral cancer treatment initiation timing since diagnosis, including radiation initiation as well as first surgery of the primary site. It also shortened the length of stay in the hospital post-surgery. The second essay assesses the value of electronic medical records from Indiana health information exchange (IHIE) and survey self-reports of Indiana residents seen at Indiana University Health in measuring population-based cancer screening for colorectal, cervical, and breast cancer. Between the two measures of screening, the study examines association using Spearman’s rank correlation and concordance using Percent Agreement and Gwet’s Agreement Coefficient. Health information exchange and self-reports, both provided unique information in measuring cancer screening, and the most robust measurement approach entails collecting screening information from both HIE and patient self-report. In this study, we find evidence of measurement error in self-reports in terms of reporting bias. The majority of the publicly available datasets collect information on cancer screening behavior through patient interviews which are self-reported and may suffer from potential measurement errors. The third essay uses a nationwide population-based database and examines the true, unbiased impact of Medicaid expansion on cancer screening for breast, colorectal, cervical, and prostate cancers after correcting for any bias due to possible misclassification of the self-reported screening status. This study conducts a modified two-way fixed effects probit model under a difference-in-difference estimation setting to identify and correct the errors in the self-reports and estimate the unbiased program effect which shows positive impact on cancer screening with increased effect sizes.

Thesis advisor: Abhishek Chatterjee / Directed evolution is a powerful technique to expand chemical space in biological systems. In particular, this method has been used to develop cellular machinery to enable genetic code expansion (GCE), the incorporation of unnatural amino acids (UAAs) into proteins during the translation process. GCE relies on evolving an aminoacyl tRNA synthetase (aaRS) and tRNA pair from a different domain of life to incorporate a UAA into proteins in their new host, as these evolutionarily distant pairs are less likely to be cross-reactive with host pairs. The aaRS and tRNA must meet a number of conditions to be useful for GCE: the pair must be orthogonal (non-cross-reactive) to the host’s native aaRS/tRNA pairs in order to ensure site-specific UAA incorporation; the aaRS must have an active site suited to accept the shape of the UAA; and the tRNA must cooperate with the host ribosome, elongation and release factors, and other translational machinery to efficiently incorporate the UAA into the protein. Numerous aaRS/tRNA pairs have been evolved to allow incorporation of diverse UAAs in bacteria due to the tractable nature of these organisms for directed evolution experiments. While an aaRS evolved in bacteria to charge a novel UAA can be used in eukaryotes, tRNAs cannot be evolved for GCE in bacteria and then used in eukaryotes because they will not have evolved in the presence of the correct translational machinery. It is necessary to evolve tRNAs directly in their host cells. Unfortunately for researchers working on GCE in mammalian cells, it is difficult to perform directed evolution on small gene products in these hosts. Transformation efficiency in mammalian cells is poor, and transient transfection yields heterogeneous DNA distribution to target cells, making selection based on performance of individual library members impossible. Viruses are an ideal DNA delivery vector for mammalian cells, as production of recombinant viruses allows control over library member generation, and viruses can be delivered with exquisite copy number control. The Chatterjee lab recently developed a platform, Virus-Assisted Directed Evolution of tRNAs (VADER), using adeno-associated virus (AAV) to evolve tRNAs for GCE directly in mammalian cells. While VADER is the first directed evolution platform that allows the evolution of small gene products in mammalian cells, its efficiency is limited by its continued reliance on transient transfection to deliver non-library DNA that is necessary for the production of rAAV. To overcome this limitation, baculovirus delivery vectors were developed to boost DNA delivery and AAV capsid production to improve virus production efficiency during selections. VADER allows the evolution of tRNAs to incorporate certain UAAs, but the technique relies on installing a UAA into the AAV capsid, which is sensitive to disruption caused by slight modifications in structure. To expand the scope of VADER to evolve tRNAs for UAAs that cannot be incorporated into the AAV capsid, an alternate selection handle (Assembly Activating Protein, or AAP) was deleted from the genome and provided in trans to incorporate 5-hydroxytryptophan (5HTP). Incorporating the UAA into this flexible protein allows UAA-dependent production of AAV and expands the scope of tRNAs that can be evolved in mammalian cells. / Thesis (PhD) — Boston College, 2021. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Chemistry.

No description available.

No description available.

No description available.

No description available.

No description available.