Genetics and Genomics Commons^™

Privacy-Preserving Federated Learning With Optimized Ensemble Weighting And Knowledge Distillation For Covid-19 Detection From Non-Iid Medical Imaging Data, Richard Annan, Hong Qin, Robert Newman, Madhuri Siddula, Letu Qingge

Computer Science Faculty Publications

Medical imaging enables rapid and accurate diagnosis of COVID-19, with CT scans proving especially effective. However, data privacy concerns limit collaborative model development across hospitals. To address this issue, we introduce a novel federated learning framework. It is referred to as Independent Knowledge Distillation with post-Ensemble Federated Learning (IKDEFL). Differential Privacy (DP) is integrated into the framework to improve privacy guarantees. Three DP mechanisms are evaluated. These include Fixed Gaussian, Gaussian Adaptive, and Tree Adaptive. The evaluation has been conducted on heterogeneous and Non-Independent and Identically Distributed (Non-IID) datasets. These datasets reflect real-world hospital scenarios. Results show that IKDEFL significantly …

Explainable Convolutional Neural Network Model Provides An Alternative Genome-Wide Association Perspective On Mutations In Sars-Cov-2, Parisa C. Hatami, Richard Annan, Luis Miranda, Jane L. Gorman, Mengjun Xie, Letu Qingge, Hong Qin

Computer Science Faculty Publications

Identifying informative genomic features in SARS-CoV-2 can help clarify patterns of viral evolution. In this study, we developed an explainable convolutional neural network (CNN) model to classify SARS-CoV-2 genomic sequences into the WHO-designated Variants of Concern (VOCs), Alpha, Beta, Gamma, Delta, and Omicron. Using a balanced dataset of genomes, the classification CNN achieved 99.96% accuracy on the held-out test set. To interpret the model’s predictions, we applied SHapley Additive exPlanations (SHAP) to estimate the contribution of each nucleotide position to VOC-label prediction and compared aggregated attributions with a chi-square GWAS baseline applied to the same categorical labels. SHAP prioritized several …

Bio-Orthogonal Chemistry-Based Strategy To Turn-On And Turn-Off Crispr-Cas9 Gene Editing In Solution And In Live Cells, Bhoomika Pandit

Electronic Theses & Dissertations (2024 - present)

The CRISPR–Cas9 system is a widely popular tool for genome engineering. There is a strong interest in developing tools for temporal control of CRISPR-Cas9 activity to address some of the challenges and to broaden the scope of potential applications. In this thesis I describe two biorthogonal based approaches to either Turn-ON and Turn-OFF CRISPR-Cas9 gene editing.

In first project work I describe a bio-orthogonal chemistry-based approach to Turn-ON Cas9 nuclease activity with temporal precision. We report a TCO-acylimidazole reagent that acylates 2′-OH groups of RNA. Poly-acylation (“cloaking”) of RNA was optimized in vitro using a model 18-nt oligonucleotide, as well …

Physiologic Markers Of Mortality In Acute Valve Syndrome: An Ischemic Physiology Score Stratifies Patient Risk, Omar Saleh, Nicholas J. Valle, Israa Saleh, Raymond Benza, Deepak R. Talreja, Matthew R. Summers

Department of Medicine Faculty Publications

Background: Acute valve syndrome (AVS) represents a high-risk phenotype of advanced valvular disease with largely uncharacterized risk heterogeneity. We aimed to validate an expanded AVS definition and develop a physiology-dependent risk stratification tool using available clinical markers.

Methods: This retrospective study analyzed 2380 patients undergoing aortic valve replacement for severe aortic stenosis and classified them as AVS (n = 1556) or progressive valvular disease (n = 824). The primary outcome was 1-year all-cause mortality. An L2-regularized logistic regression model was developed to predict 1-year mortality using admission laboratory markers and comorbidity burden. Model performance was assessed using nested cross-validation with …

Phylogeography, Fungal Host Specificity, And Fitness In The Corallorhiza Maculata Species Complex, Hana Thixton-Nolan

Graduate Theses, Dissertations, and Problem Reports (ETD)

Mycoheterotrophic plants exploit fungi for most or all nutritional needs, providing a compelling system to address the evolutionary consequences of genome reduction, population-level variation, and germination requirements. I am using the widespread Corallorhiza maculata species complex to address questions of genomic variation, fungal host specificity, morphological variation, and germination requirements across the broad geographic range from Mexico to northern North America. This complex contains C. bulbosa, C. macrantha, and C. mertensiana, the latter with three varieties (maculata, occidentalis, and mexicana). These taxa vary in flowering time, floral morphology, and fungal associations. However, the degree …

Genetics Of Mycangial Development And Host Colonization Within The Ambrosia Symbiosis, Mckeon James Laws

Graduate Theses, Dissertations, and Problem Reports (ETD)

Ambrosia beetles, a polyphyletic group of weevils, utilize pouch-like structures at various locations on the body to transport for their sole food source, their fungal partner. Ambrosia fungi are known to originate from multiple unrelated fungal lineages with no consensus of traits required to become a fungal cultivar. While the biology and ecology of both partners is in many cases well understood, the genetic underpinnings of this fascinating mutualism remain a mystery. The aim of this dissertation is to add to the understanding of the ambrosia symbiosis by providing insights into its molecular and genetic underpinnings through RNA sequencing data, …