Data Science Commons^™

Intelligent Resource Prediction For Hpc And Scientific Workflows, Benjamin Shealy

All Dissertations

Scientific workflows and high-performance computing (HPC) platforms are critically important to modern scientific research. In order to perform scientific experiments at scale, domain scientists must have knowledge and expertise in software and hardware systems that are highly complex and rapidly evolving. While computational expertise will be essential for domain scientists going forward, any tools or practices that reduce this burden for domain scientists will greatly increase the rate of scientific discoveries. One challenge that exists for domain scientists today is knowing the resource usage patterns of an application for the purpose of resource provisioning. A tool that accurately estimates these …

Characterizing Long Covid: Deep Phenotype Of A Complex Condition, Rachel R. Deer, Madeline A. Rock, Nicole Vasilevsky, Leigh Carmody, Halie Rando, Alfred J. Anzalone, Marc D. Basson, Tellen D. Bennett, Timothy Bergquist, Eilis A. Boudreau, Carolyn T. Bramante, James Brian Byrd, Tiffany J. Callahan, Lauren E. Chan, Haitao Chu, Christopher G. Chute, Ben D. Coleman, Hannah E. Davis, Joel Gagnier, Casey S. Greene, Ramakanth Kavuluru

Institute for Biomedical Informatics Faculty Publications

BACKGROUND: Numerous publications describe the clinical manifestations of post-acute sequelae of SARS-CoV-2 (PASC or "long COVID"), but they are difficult to integrate because of heterogeneous methods and the lack of a standard for denoting the many phenotypic manifestations. Patient-led studies are of particular importance for understanding the natural history of COVID-19, but integration is hampered because they often use different terms to describe the same symptom or condition. This significant disparity in patient versus clinical characterization motivated the proposed ontological approach to specifying manifestations, which will improve capture and integration of future long COVID studies.

METHODS: The Human Phenotype Ontology …

Comprehensive Characterization Of Covid-19 Patients With Repeatedly Positive Sars-Cov-2 Tests Using A Large Us Electronic Health Record Database, Xiao Dong, Yujia Zhou, Xiao-Ou Shu, Elmer V Bernstam, Rebecca Stern, David M Aronoff, Hua Xu, Loren Lipworth

Faculty, Staff and Student Publications

In the absence of genome sequencing, two positive molecular tests for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) separated by negative tests, prolonged time, and symptom resolution remain the best surrogate measure of possible reinfection. Using a large electronic health record database, we characterized clinical and testing data for 23 patients with repeatedly positive SARS-CoV-2 PCR test results ≥60 days apart, separated by ≥2 consecutive negative test results. The prevalence of chronic medical conditions, symptoms, and severe outcomes related to coronavirus disease 19 (COVID-19) illness were ascertained. The median age of patients was 64.5 years, 40% were Black, and 39% …

Digital Technology Needs In Maternal Mental Health: A Qualitative Inquiry, Alexandra Zingg, Laura Carter, Deevakar Rogith, Amy Franklin, Sudhakar Selvaraj, Jerrie Refuerzo, Sahiti Myneni

Faculty, Staff and Student Publications

Digital technologies offer many opportunities to improve mental healthcare management for women seeking pre- and-postnatal care. They provide a discrete, practical medium that is well-suited for the sensitive nature of mental health. Women who are more prone to experiencing peripartum depression (PPD), such as those of low-socioeconomic background or in high-risk pregnancies, can benefit the most from such technologies. However, current digital interventions directed towards this population provide suboptimal support, and their responsiveness to end user needs is quite limited. Our objective is to understand the digital terrain of information needs for low-socioeconomic status women with high-risk pregnancies, specifically within …

Med-Bert: Pretrained Contextualized Embeddings On Large-Scale Structured Electronic Health Records For Disease Prediction, Laila Rasmy, Yang Xiang, Ziqian Xie, Cui Tao, Degui Zhi

Faculty, Staff and Student Publications

Deep learning (DL)-based predictive models from electronic health records (EHRs) deliver impressive performance in many clinical tasks. Large training cohorts, however, are often required by these models to achieve high accuracy, hindering the adoption of DL-based models in scenarios with limited training data. Recently, bidirectional encoder representations from transformers (BERT) and related models have achieved tremendous successes in the natural language processing domain. The pretraining of BERT on a very large training corpus generates contextualized embeddings that can boost the performance of models trained on smaller datasets. Inspired by BERT, we propose Med-BERT, which adapts the BERT framework originally developed …

Unsupervised And Supervised Learning For Rna-Protein Interactions And Annotations, Kateland Sipe

Honors Projects

This project analyzed the base and amino acid interactions and annotations through the use of unsupervised and supervised learning techniques. For unsupervised learning, clustering found the data was not able to be distinguished into clear groups which matched the original annotations through kmeans clustering and hierarchical clustering. For supervised learning, the use of random forest, glmnet, and deep learning neural networks were successful in creating accurate predictions. However, machine learning likely will not be able to replace the original complex program, but could be used for possible simplification.

Generalized And Transferable Patient Language Representation For Phenotyping With Limited Data, Yuqi Si, Elmer V Bernstam, Kirk Roberts

Faculty, Staff and Student Publications

The paradigm of representation learning through transfer learning has the potential to greatly enhance clinical natural language processing. In this work, we propose a multi-task pre-training and fine-tuning approach for learning generalized and transferable patient representations from medical language. The model is first pre-trained with different but related high-prevalence phenotypes and further fine-tuned on downstream target tasks. Our main contribution focuses on the impact this technique can have on low-prevalence phenotypes, a challenging task due to the dearth of data. We validate the representation from pre-training, and fine-tune the multi-task pre-trained models on low-prevalence phenotypes including 38 circulatory diseases, 23 …

Deep Learning For Multi-Tissue Cancer Classification Of Gene Expressions, Tarek Khorshed

Theses and Dissertations

We contribute in saving the lives of cancer patients through early detection and diagnosis, since one of the major challenges in cancer treatment is that patients are diagnosed at very late stages when appropriate medical interventions become less effective and full curative treatment is no longer achievable. Cancer classification using gene expressions is extremely challenging given the complexity and high dimensionality of the data. Current classification methods typically rely on samples collected from a single tissue type and perform a prerequisite of gene feature selection to avoid processing the full set of genes. These methods fall short in taking advantage …

Exchanges In A Virtual Environment For Diabetes Self-Management Education And Support: Social Network Analysis, Carlos A Pérez-Aldana, Allison A Lewinski, Constance M Johnson, Allison A Vorderstrasse, Sahiti Myneni

Faculty, Staff and Student Publications

BACKGROUND: Diabetes remains a major health problem in the United States, affecting an estimated 10.5% of the population. Diabetes self-management interventions improve diabetes knowledge, self-management behaviors, and clinical outcomes. Widespread internet connectivity facilitates the use of eHealth interventions, which positively impacts knowledge, social support, and clinical and behavioral outcomes. In particular, diabetes interventions based on virtual environments have the potential to improve diabetes self-efficacy and support, while being highly feasible and usable. However, little is known about the patterns of social interactions and support taking place within type 2 diabetes-specific virtual communities.

OBJECTIVE: The objective of this study was to …

Red Drum And Spotted Seatrout Live-Release Tournament Mortality And Dispersal, T. Reid Nelson, Crystal Hightower, Sean P. Powers

University Faculty and Staff Publications

Although catch-and-release fishing tournaments undoubtedly reduce mortality of target species, postrelease mortality and fish stockpiling at release sites remain common concerns related to these tournaments. The impacts of liverelease tournaments on freshwater species have been widely studied. However, research on estuarine sport fishes is lacking even though catch-and-release tournaments targeting these species are prevalent and popular recreational fisheries exist. Therefore, we estimated the post-weigh-in mortality and dispersal of Red Drum Sciaenops ocellatus and Spotted Seatrout Cynoscion nebulosus released from the 2016–2018 Alabama Deep Sea Fishing Rodeo live-weigh-in categories using acoustic telemetry. To concurrently estimate overall post-weigh-in mortality and dispersal, we …

Machine Learning And Bioinformatic Insights Into Key Enzymes For A Bio-Based Circular Economy, Japheth E. Gado

Theses and Dissertations--Chemical and Materials Engineering

The world is presently faced with a sustainability crisis; it is becoming increasingly difficult to meet the energy and material needs of a growing global population without depleting and polluting our planet. Greenhouse gases released from the continuous combustion of fossil fuels engender accelerated climate change, and plastic waste accumulates in the environment. There is need for a circular economy, where energy and materials are renewably derived from waste items, rather than by consuming limited resources. Deconstruction of the recalcitrant linkages in natural and synthetic polymers is crucial for a circular economy, as deconstructed monomers can be used to manufacture …

A Multi-Resolution Graph Convolution Network For Contiguous Epitope Prediction, Lisa Oh

Dartmouth College Master’s Theses

Computational methods for predicting binding interfaces between antigens and antibodies (epitopes and paratopes) are faster and cheaper than traditional experimental structure determination methods. A sufficiently reliable computational predictor that could scale to large sets of available antibody sequence data could thus inform and expedite many biomedical pursuits, such as better understanding immune responses to vaccination and natural infection and developing better drugs and vaccines. However, current state-of-the-art predictors produce discontiguous predictions, e.g., predicting the epitope in many different spots on an antigen, even though in reality they typically comprise a single localized region. We seek to produce contiguous predicted epitopes, …

A Comparison Of Exhaustive And Non-Lattice-Based Methods For Auditing Hierarchical Relations In Gene Ontology, Rashmie Abeysinghe, Fengbo Zheng, Licong Cui

Faculty, Staff and Student Publications

Uncovering and fixing errors in biomedical terminologies is essential so that they provide accurate knowledge to downstream applications that rely on them. Non-lattice-based methods have been applied to identify various kinds of inconsistencies in different biomedical terminologies. In previous work, we have introduced two inference-based approaches that were applied in an exhaustive manner to audit hierarchical relations in the Gene Ontology: (1) Lexical-based inference framework, and (2) Subsumption-based sub-term inference framework. However, it is unclear how effective these exhaustive approaches perform compared with their corresponding non-lattice-based approaches. Therefore, in this paper, we implement the non-lattice versions of these two exhaustive …

Identifying Sleep-Related Factors Associated With Cognitive Function In A Hispanics/Latinos Cohort: A Dual Random Forest Approach, Li Xiaojin, Cui Licong, Wang Fei, Paul E Schulz, Guo-Qiang Zhang

Faculty, Staff and Student Publications

Disordered sleep is associated with poor cognitive function and cognitive decline. However, little is known regarding the association of sleep-related factors with cognitive function in underrepresented cohorts such as the Hispanic/Latino population. Leveraging the National Sleep Research Resource, one of the most comprehensive collections of sleep studies, we identified a Hispanic/Latino cohort of 1,031 lower cognitive function cases and 2,062 normal controls. We developed a novel dual random forest (DRF) approach to discriminate cases against controls for estimating the potential impact of sleep-related variables related to the decline of cognitive function. Several important sleep-related factors were identified which may be …

Ensemble Protein Inference Evaluation, Kyle Lee Lucke

Graduate Student Theses, Dissertations, & Professional Papers

The Protein inference problem is becoming an increasingly important tool that aids in the characterization of complex proteomes and analysis of complex protein samples. In bottom-up shotgun proteomics experiments the metrics for evaluation (like AUC and calibration error) are based on an often imperfect target-decoy database. These metrics make the inherent assumption that all of the proteins in the target set are present in the sample being analyzed. In general, this is not the case, they are typically a mix of present and absent proteins. To objectively evaluate inference methods, protein standard datasets are used. These datasets are special in …

Stochastic Modeling Of Ovarian Follicle Growth In Adult Female Rats, Zhaozhi Li

Annual Symposium on Biomathematics and Ecology Education and Research

No abstract provided.

Representation Of Ehr Data For Predictive Modeling: A Comparison Between Umls And Other Terminologies, Laila Rasmy, Firat Tiryaki, Yujia Zhou, Yang Xiang, Cui Tao, Hua Xu, Degui Zhi

Faculty, Staff and Student Publications

OBJECTIVE: Predictive disease modeling using electronic health record data is a growing field. Although clinical data in their raw form can be used directly for predictive modeling, it is a common practice to map data to standard terminologies to facilitate data aggregation and reuse. There is, however, a lack of systematic investigation of how different representations could affect the performance of predictive models, especially in the context of machine learning and deep learning.

MATERIALS AND METHODS: We projected the input diagnoses data in the Cerner HealthFacts database to Unified Medical Language System (UMLS) and 5 other terminologies, including CCS, CCSR, …

Machine Learning Applications For Drug Repurposing, Hansaim Lim

Dissertations, Theses, and Capstone Projects

The cost of bringing a drug to market is astounding and the failure rate is intimidating. Drug discovery has been of limited success under the conventional reductionist model of one-drug-one-gene-one-disease paradigm, where a single disease-associated gene is identified and a molecular binder to the specific target is subsequently designed. Under the simplistic paradigm of drug discovery, a drug molecule is assumed to interact only with the intended on-target. However, small molecular drugs often interact with multiple targets, and those off-target interactions are not considered under the conventional paradigm. As a result, drug-induced side effects and adverse reactions are often neglected …

Understanding Spatial Language In Radiology: Representation Framework, Annotation, And Spatial Relation Extraction From Chest X-Ray Reports Using Deep Learning, Surabhi Datta, Yuqi Si, Laritza Rodriguez, Sonya E Shooshan, Dina Demner-Fushman, Kirk Roberts

Faculty, Staff and Student Publications

Radiology reports contain a radiologist's interpretations of images, and these images frequently describe spatial relations. Important radiographic findings are mostly described in reference to an anatomical location through spatial prepositions. Such spatial relationships are also linked to various differential diagnoses and often described through uncertainty phrases. Structured representation of this clinically significant spatial information has the potential to be used in a variety of downstream clinical informatics applications. Our focus is to extract these spatial representations from the reports. For this, we first define a representation framework based on the Spatial Role Labeling (SpRL) scheme, which we refer to as …

Covid-19 Testnorm: A Tool To Normalize Covid-19 Testing Names To Loinc Codes, Xiao Dong, Jianfu Li, Ekin Soysal, Jiang Bian, Scott L Duvall, Elizabeth Hanchrow, Hongfang Liu, Kristine E Lynch, Michael Matheny, Karthik Natarajan, Lucila Ohno-Machado, Serguei Pakhomov, Ruth Madeleine Reeves, Amy M Sitapati, Swapna Abhyankar, Theresa Cullen, Jami Deckard, Xiaoqian Jiang, Robert Murphy, Hua Xu

Faculty, Staff and Student Publications

Large observational data networks that leverage routine clinical practice data in electronic health records (EHRs) are critical resources for research on coronavirus disease 2019 (COVID-19). Data normalization is a key challenge for the secondary use of EHRs for COVID-19 research across institutions. In this study, we addressed the challenge of automating the normalization of COVID-19 diagnostic tests, which are critical data elements, but for which controlled terminology terms were published after clinical implementation. We developed a simple but effective rule-based tool called COVID-19 TestNorm to automatically normalize local COVID-19 testing names to standard LOINC (Logical Observation Identifiers Names and Codes) …

Two (Or More) Viruses In One Bat: A Systematic Quantitative Literature Review Of Viral Coinfection In Bats, Eli J. Kaufman

Independent Study Project (ISP) Collection

Viral coinfection is an important topic in pathogen dynamics, and can increase viral shedding and change disease outcomes. As bats are carriers of important zoonoses, such as the SARS coronaviruses, rabies, and other deadly viruses, knowing more about their coinfection dynamics is important. This quantitative systematic literature review sought to show how many papers reported bat viral coinfections, and created three databases. The first database, the SQLR database was based on searches for coinfections. The second database, the Astrovirus database was to determine how much of the literature was being missed by examining a single viral family more in depth …

Deep Learning In Clinical Natural Language Processing: A Methodical Review, Stephen Wu, Kirk Roberts, Surabhi Datta, Jingcheng Du, Zongcheng Ji, Yuqi Si, Sarvesh Soni, Qiong Wang, Qiang Wei, Yang Xiang, Bo Zhao, Hua Xu

Faculty, Staff and Student Publications

OBJECTIVE: This article methodically reviews the literature on deep learning (DL) for natural language processing (NLP) in the clinical domain, providing quantitative analysis to answer 3 research questions concerning methods, scope, and context of current research.

MATERIALS AND METHODS: We searched MEDLINE, EMBASE, Scopus, the Association for Computing Machinery Digital Library, and the Association for Computational Linguistics Anthology for articles using DL-based approaches to NLP problems in electronic health records. After screening 1,737 articles, we collected data on 25 variables across 212 papers.

RESULTS: DL in clinical NLP publications more than doubled each year, through 2018. Recurrent neural networks (60.8%) …

Prospects And Challenges Of Population Health With Online And Other Big Data In Africa; Understanding The Link To Improving Healthcare Service Delivery, Rowland Edet, Bolarinwa Afolabi

Department of Sociology: Faculty Publications

Big data analytics offers promises to many health care service challenges and can provide answers to many population health issues. Big data is having a positive impact in almost every sphere of life in more advanced world while developing countries are striving to meet up. Even though healthcare systems in the developed world are recording some breakthroughs due to the application of big data, it is important to research the impact of big data in developing regions of the world, such as Africa and identify its peculiar needs. The purpose of this review was to summarize the challenges faced by …

Digilego For Peripartum Depression: A Novel Patient-Facing Digital Health Instantiation, J Rodin, C Timko, S Harris

Faculty, Staff and Student Publications

Digital health technologies offer unique opportunities to improve health outcomes for mental health conditions such as peripartum depression (PPD), a disorder that affects approximately 10-15% of women in the U.S. every year. In this paper, we present the adaption of a digital technology development framework, Digilego, in the context of PPD. Methods include mapping of the Behavior Intervention Technology (BIT) model and the Patient Engagement Framework (PEF) to translate patient needs captured through focus groups. This informs formative development and implementation of digital health features for optimal patient engagement in PPD screening and management. Results show an array ofPPD-specific Digilego …

Causal Discovery In Radiographic Markers Of Knee Osteoarthritis And Prediction For Knee Osteoarthritis Severity With Attention-Long Short-Term Memory, Yanfei Wang, Lei You, Jacqueline Chyr, Lan Lan, Weiling Zhao, Yujia Zhou, Hua Xu, Philip Noble, Xiaobo Zhou

Faculty, Staff and Student Publications

The goal of this study is to build a prognostic model to predict the severity of radiographic knee osteoarthritis (KOA) and to identify long-term disease progression risk factors for early intervention and treatment. We designed a long short-term memory (LSTM) model with an attention mechanism to predict Kellgren/Lawrence (KL) grade for knee osteoarthritis patients. The attention scores reveal a time-associated impact of different variables on KL grades. We also employed a fast causal inference (FCI) algorithm to estimate the causal relation of key variables, which will aid in clinical interpretability. Based on the clinical information of current visits, we accurately …

Enhancing Clinical Concept Extraction With Contextual Embeddings, Yuqi Si, Jingqi Wang, Hua Xu, Kirk Roberts

Faculty, Staff and Student Publications

OBJECTIVE: Neural network-based representations ("embeddings") have dramatically advanced natural language processing (NLP) tasks, including clinical NLP tasks such as concept extraction. Recently, however, more advanced embedding methods and representations (eg, ELMo, BERT) have further pushed the state of the art in NLP, yet there are no common best practices for how to integrate these representations into clinical tasks. The purpose of this study, then, is to explore the space of possible options in utilizing these new models for clinical concept extraction, including comparing these to traditional word embedding methods (word2vec, GloVe, fastText).

MATERIALS AND METHODS: Both off-the-shelf, open-domain embeddings and …

Precision Agriculture Gis Technologies For Mississippi, 1st. Edition, Amelia A.A. Fox

CALS Publications

Precision agriculture is meant to improve on-farm efficiency in hopes of ultimately increasing profitability while also protecting the environment. However, this difficult process almost always includes the proper management and interpretation of data. Therefore, it is imperative that those individuals involved in making such decisions are educated on these processes. In a data-driven world, this textbook is a great resource for those wanting to learn how to utilize their data in hopes of making better informed on-farm decisions.

Deep Patient Representation Of Clinical Notes Via Multi-Task Learning For Mortality Prediction, Yuqi Si, Kirk Roberts

Faculty, Staff and Student Publications

We propose a deep learning-based multi-task learning (MTL) architecture focusing on patient mortality predictions from clinical notes. The MTL framework enables the model to learn a patient representation that generalizes to a variety of clinical prediction tasks. Moreover, we demonstrate how MTL enables small but consistent gains on a single classification task (e.g., in-hospital mortality prediction) simply by incorporating related tasks (e.g., 30-day and 1-year mortality prediction) into the MTL framework. To accomplish this, we utilize a multi-level Convolutional Neural Network (CNN) associated with a MTL loss component. The model is evaluated with 3, 5, and 20 tasks and is …

Effects Of A Community Population Health Initiative On Blood Pressure Control In Latinos, James R Langabeer, Timothy D Henry, Carlos Perez Aldana, Larissa Deluna, Nora Silva, Tiffany Champagne-Langabeer

Faculty, Staff and Student Publications

Background Hypertension remains one of the most important, modifiable cardiovascular risk factors. Yet, the largest minority ethnic group (Hispanics/Latinos) often have different health outcomes and behavior, making hypertension management more difficult. We explored the effects of an American Heart Association-sponsored population health intervention aimed at modifying behavior of Latinos living in Texas. Methods and Results We enrolled 8071 patients, and 5714 (65.7%) completed the 90-day program (58.5 years ±11.7; 59% female) from July 2016 to June 2018. Navigators identified patients with risk factors; initial and final blood pressure ( BP ) readings were performed in the physician's office; and interim …

Estimating Exploitation Rates In The Alabama Red Snapper Fishery Using A High-Reward Tag–Recapture Approach, Dana K. Sackett, Mattgew Catalano, J. Marcus Drymon, Sean P. Powers, Mark Albins

University Faculty and Staff Publications

Accurate estimates of exploitation are essential to managing an exploited fishery. However, these estimates are often dependent on the area and vulnerable sizes of fish considered in a study. High-reward tagging studies offer a simple and direct approach to estimating exploitation rates at these various scales and in examining how model parameters impact exploitation rate estimates. These methods can ultimately provide a better understanding of the spatial dynamics of exploitation at smaller local and regional scales within a fishery—a measure often needed for more site-attached species, such as the Red Snapper Lutjanus campechanus. We used this approach to tag 724 …