Welcome to the BiosBeat communication tool for the Department of Biostatistics at the UNC Gillings School of Global Public Health! Here you will find a collection of the latest department news, special features, dates to save, and much more. So, read on, enjoy, and be sure to regularly check back for updates!


Message from our Biostatistics Leadership – COVID19

We hope you are staying safe and healthy during this time. We encourage you to continue taking all the preventative measures recommended by UNC and the CDC.

Please visit the School of Public Health’s official Coronavirus Information Portal for the most recent information and updates: https://sph.unc.edu/global-health/2019-coronavirus-info-portal/




Dr. Cicely Mitchell

Dr. Cicely Mitchell


Cicely Mitchell was the Gillings School commencement ceremony speaker in May. Mitchell received both her Master of Science and Doctor of Public Health degrees in biostatistics from UNC. Read more about the occasion and our alum here. Find the speech video here.



Donald Fejfar, Biostatistics undergraduate student

Starting freshman year at the Gillings School of Global Public Health, Biostatistics undergraduate Donald Fejfar (BSPh 2021) has been involved in global health research projects. In his first semester, he started at Gillings Water Institute, collaborating on several papers about water, sanitation, and hygiene throughout his college experience. This work led to the Galapagos Islands project (read more here), where Fejfar led an undergraduate team to conduct research on water and food insecurity on Isabela Island and its relation to the impact on the health of its citizens.

Once the COVID-19 pandemic hit, he began interning with Harvard and the Partners in Global Health Research Core doing biostatistical work. His work included cleaning and processing data from Partners in Health care sites in eight countries to use the data in Syndromic Surveillance. Syndromic Surveillance is a monitoring activity where screening is done for indicators of COVID-19 infection at clinics to see if infection rates are spiking compared to predictions using models and data from the baseline period. At the beginning of the pandemic, many countries did not have the capacity to test for COVID-19 directly. After processing the data, Fejfar also analyzed the data to look into other factors such as immunization rates and maternal health indicators during COVID-19.

Fejfar finished his senior honors thesis this year that focused on the relationship between domestic cooking fuels used in China and host plasma metabolites and gut microbiota. Recently, there has been a surge in the field of pollution research with a focus on indoor air pollution and its effects on human health. One major element that researchers evaluate and compare is the health of people who use “clean” energy fuels while cooking versus the people who use “dirty” cooking fuels. Indoor air pollution can be linked to several different health conditions, such as diabetes, chronic heart disease, and can put people at a higher risk for premature death. However, the pathway linking the two is unclear and is still under question. With the help of his mentors Gillings Professors Annie Green Howard and Penny Gordon-Larsen, Fejfar’s research attempted to start exploring the relationship between indoor pollution and health issues. “[This] was preliminary research meant to set the stage for more complex analyses to really dissect these relationships,” Fejfar said, “my project was to start piecing together the associations between and measurements of each variable. I worked with my mentors to follow their lead and do plenty of descriptive and random forest analyses on the Chinese Health and Nutrition Survey data.”

Fejfar’s Gilling’s School education has been invaluable throughout his time in the research field. He said, “BIOS is a great major because the skills are super flexible and you can take them to whatever projects you want to work on, [from] fieldwork in the Galapagos [to] remote work on massive datasets from China or Lesotho, you name it!” He also said that through his time at UNC, he learned how much room for growth truly exists, especially in biostatistics and global health, and that there is always room to learn more. After graduating, Fejfar will be joining the Harvard and Partners in Global Public Health Research Core team full-time as a research analyst and plans to apply to medical school.



Machine learning, in particular deep learning, has grown dramatically in popularity over recent years. There are many applications of deep learning that can be used to solve problems within the biomedical sciences. However, the greater prevalence and complexity of missing data in biomedical datasets presents significant challenges for deep learning methods. While rich literature exists for the treatment of missing data in traditional statistic models, formal methods for missing data in deep learning is lacking. This is the area where Gillings Biostatistics PhD student David Lim’s research has been focused. David is involved in a project known as Non-Ignorably Missing Importance-Weighted Autoencoder, or NIMIWAE, along with Biostatistics Professors Rashid and Ibrahim.

David Lim is pursuing his PhD in biostatistics at the Gillings School

David double majored in Applied Mathematics and Physics at UCLA and sought an opportunity to apply the theoretical concepts that he had learned in his undergraduate studies into real-life applications through a PhD in Biostatistics at the Gillings School.

“I initially looked at deep learning methods in single-cell RNA-seq data, and thought a very nice application of deep learning could be in missing data, where there is a need for a method to properly handle different mechanisms of missingness in increasingly larger and higher-dimensional datasets. My previous research in clustering bulk RNA-seq gene expression data was rather different, although one can think about cluster membership in that setting as a latent variable, just as we treat the missing data in NIMIWAE as a latent variable.” David shared.

The way that the NIMIWAE software works is that it takes a data matrix with incomplete entries and trains a deep learning neural network that learns a probabilistic model of the data and imputes missing entries.

“The user can specify the assumption of whether the mechanism of missingness is ignorable, and, for a non-ignorable model, a missingness model is learned using a neural network. The user can specify select covariates for the missingness model, or choose to include all features in the original data as covariates. Additionally, missing entries are sampled from an approximate posterior of the variable denoting the missing values in the data.” David explains.

David’s primary role was to program the R package to perform the deep learning heavy-lifting in a reticulated Python backend. He mentions that his education in missing data mechanisms and maximum likelihood methods was particularly valuable in this project.

David and the writing team are currently in the process of revising the paper, and they plan to include functionality of NIMIWAE to handle mixed data types.

“We are also concurrently working on a sequel of this project (called “dlglm”) to perform supervised learning/prediction within a generalized linear model framework using another novel deep learning architecture.” David shared.

See more about NIMIWAE.


The most recent Journal of the American Statistical Association (JASA) Theory and Methods section has a special focus on precision medicine and individualized policy discovery. This edition was guest-edited by our own Professor Michael Kosorok, and also features Biostatistics Professor Donglin Zeng as an associate editor. Some of the contents of this issue include papers from Gillings School faculty members, including Biostatistics Professors Michael Kosorok, Donglin Zeng, and Yufeng Liu.

The basic concept behind precision medicine is focusing on data and clinical knowledge to personalize medical treatment for each individual patient. Since each patient’s health history and health status is unique, the treatment they are given must be tailored to fit their individual needs. The data that describes a patient’s heath can be complex and irregular. The challenges that come with trying to synthesize such data in order to determine the best plan of treatment for patients are numerous, many of which are touched on in this JASA issue.

This issue emphasizes the use of data to improve decision-making. This includes the development of robust and efficient methods for estimation of personalized intervention recommendation systems in addition to the generation of new knowledge about a particular decision process under study. While the findings in this issue primarily focus on biomedical applications, the ideas discussed have the potential to be relevant to a wide range of areas, including business, engineering, public policy, and education.

Professor Kosorok, Guest Editor JASA

The Gillings School holds a central role in the field of precision medicine, with several faculty members being involved in innovative, cutting-edge research surrounding the topic. One of the papers featured in the issue, “Learning Individualized Treatment Rules for Multiple-Domain Latent Outcomes,” is co-authored by Biostatistics Professor Donglin Zeng. In this paper, the authors use a restricted Boltzmann machine to develop methods for estimating individualized treatment rules for optimizing latent outcomes, such as latent mental health status based on multiple-domain psychological or clinical symptoms.

Professor Kosorok said, “Faculty and students in UNC Biostatistics, along with many other participants in the Gillings School and elsewhere on UNC campus, are providing national and international leadership on cutting-edge precision health at the confluence of machine learning, causal inference and human health. It is an exciting area to work in and watch develop.”

Read the full special-edition, precision medicine and individualized policy discovery edition of the JASA Theory and Methods section.


Matthew Psioda

Assistant Professor Matthew Psioda

Biostatistics Assistant Professor Matt Psioda, along with Professor Joseph G. Ibrahim, doctoral student Jiawei Xu, and collaborators at Amgen Inc, has recently published an oncology basket trial design paper, “Bayesian adaptive basket trial design using model averaging” in the journal Biostatistics that seeks to borrow information using an adaptive trial design based on Bayesian model averaging.

Oncology drug development has historically focused on the histology of cancer, as this was the key known determinant for whether a tumor would respond to a given treatment. With the emergence of genomic technologies that allow for the characterization of specific genetic mutations within a tumor, the focus has now broadened to include developing new types of targeted therapies for specific mutations. New types of trials called “basket trials” are being conducted that enroll cancer patients in cases where the tumors contain a specific mutation that an investigational treatment has been designed to target. The goal of these studies is to identify the subset of tumor types where an investigational treatment has efficacy. Frequently, Bayesian approaches are used to borrow information across tumor types to better estimate intervention effects for each of them. This is in contrast to classical approaches where data for each tumor type are essentially evaluated independent of other data.

More recently, the same team published a second paper that proposed a novel Bayesian approach to extend the application of basket trials to new settings such as the study of inflammatory diseases. For inflammatory diseases, certain proteins (e.g., cytokines such as Interleukin 17) are often implicated in the disease process for multiple diseases, and novel treatments are being developed that target them, motivating the use of basket trial designs. This paper titled “Bayesian adaptive design for concurrent trials involving biologically related diseases,” also published in the journal Biostatistics, has the potential to make early phase drug development programs that involve multiple related diseases more efficient.

Both of these projects stem from ongoing research collaborations as a part of the UNC Laboratory for Innovative Clinical Trials, which, in part, focuses on methods for trial design and analysis using Bayesian methods (e.g., power priors, response-adaptive designs).



Liz Zarzar, BSPH 2021 (2nd from left) with some of her graduating classmates

Now that graduation has happened, Liz Zarzar, a recent graduate of UNC and the Department of Biostatistics, has now had time to reflect on her past four years. The memories she made and the people she met will always make Chapel Hill home for her. Her sophomore year, she decided to declare a BSPH degree at Gillings, which is where she discovered her passion for the intersection of big data and human health.

For many students, the COVID-19 pandemic was a unique and sometimes stressful experience. The transition to remote learning was difficult for everyone and brought up many new challenges. However, as a student of Public Health, Zarzar says it was fascinating to see the public health systems at work in real-time. She says, “My classes at Gillings taught me all about public health systems and solutions, clinical trials, and designs of public health studies, sampling populations, and predictive modeling. All of this content has extreme relevance to the COVID-19 crisis, and it was engaging to have discussions about current events with top professors in the field every day in class.” She goes on to say that public health is dynamic, which is one of the reasons she loves it and is grateful to have the opportunity to address a variety of these topics in her post-graduation job.

After graduation, Zarzar will be starting a job as a Technical Consultant with SAS. She will be working in the office in Washington D.C. in the field of Public Sector Analytics (more information can be found here.) She says, “I am so excited to use the skills I’ve gained from the BSPH program to help government entities make data-driven decisions to improve the safety and well-being of citizens. By working in the public sector, I will have the opportunity to tackle a variety of projects, and some will help make evidence-based decisions and solutions for combating the impacts of the COVID-19 crisis.”

The BSPH curriculum provided Zarzar with real-world programming experience. Because of this, she now has a strong statistical background with analytical skills, but she hopes to continue to add to the analytical toolbox and be a lifelong learner in data science. “I know working at SAS will put me at the forefront of advancements in big data technology, and I am excited to help a variety of public sector departments and organizations implement SAS and digital transformation to further their missions.”



Dr. Michael Love

Assistant Professor Michael Love

Assistant Professor of Biostatistics and Genetics, Michael Love, is one of seven recipients to receive the prestigious student-nominated Teaching Excellence and Innovation Award. This award was first presented in 2012 to recognize instructors who students feel “improve the learning environment at the Gillings School by integrating new technologies, engaging students in interactive activities, employing creative assessment methods, and introducing and incorporating progressive curriculum ideas into the classroom.”

Professor Love joined the faculty in 2016 and has taught an array of students since his start at UNC. One student reflects on their experience in Love’s class, saying, “It is really a privilege to receive instruction under Dr. Love. While he is the expert in his field, he can convey the information to those of us who are relatively new to the field in a manner that makes us feel engaged and intelligent. He is enthusiastic about everything and anything he teaches, and this enthusiasm promotes discussion and diversity of ideas in the classroom. You would never know that we are attending a Zoom class – everyone actually wants to keep their cameras on during his lecture!”

Professor Love is very deserving of this award, and the Biostatistics Department is proud to have him!


Di Wu, PhD

Associate Professor Di Wu

Congratulations to Professor Di Wu on promotion to associate professor level! Wu works in the bioinformatics field and has developed statistical bioinformatics methods to handle biomedical and genomics data. Her area of interest includes understanding the relation between datasets via data integration strategy to understand the biological mechanisms related to diseases.  Wu has a joint appointment with the Dental School.




Matthew Psioda

Assistant Professor Baiming Zou


Name:  Baiming Zou

Position:  Research Assistant Professor

Time at the Gillings School: Since January 2020

What I do (and why I love it): I specialize in large electronic health record (EHR) data analysis by using machine learning methods.  EHR data reflects daily clinical practices and provides valuable resources to address many clinical problems with complicated data architectures, and machine learning algorithms are useful analytic tools to overcome these challenges.

First job or Internship I had was: Research assistant professor at the University of Florida.

Melissa Hobgood beside Rammes overlooking students at orientation day


Melissa Hobgood is known as a backbone in the biostatistics department. Her positivity encourages and motivates both her coworkers and the students that she works with daily. Because of this, Hobgood has been awarded the Employee Forum’s 2021 Hall of Famer (Staff Member of the Year). This is awarded to an individual who consistently exemplifies the University’s mission of “integrity, collaboration, respect & high-level customer service, and has been with the University for at least 5 years.”

One of her nominators and coworker, Johnathan Earnest, says, “Melissa Hobgood is quite possibly the most positive team member with whom I’ve ever worked. Despite the extremely challenging year she has faced personally, she has continued to inspire her team and students to be our best selves, to care for one another, and to model respect and integrity in all we do.”

Biostatistics Student Association (BSA)  Social Chair, Ann Marie Weideman, says “There is a famous saying by Elizabeth Gilbert: “Embrace the glorious mess that you are.” We BIOS students tend to strive for perfection; it’s in our nature. Melissa floods our inbox with encouraging emails constantly reminding us of how beautifully unbalanced we are, and many of us take advantage of her open-door policy to confess our mess. There is no one more deserving of accolades for her exceptional service to the University.”

Congratulations, Melissa!


End of semester BSA visit to the NC Botanical Gardens

The Biostatistics Student Association (BSA) kicked off summer with an excursion to the North Carolina Botanical Garden at UNC Chapel Hill. Providing a much-needed break from the flurry of Zoom calls, allowing some time without face masks, and a distraction from studying for exams. Connecting with nature after much keyboard and screen time allowed the students to unwind and catch up with friends.



This time of year is full of congratulatory messages for award winners and good wishes for our graduates embarking on new adventures! The Biostatistics Awards Ceremony was held April 29, 2021, to honor new inductees into the Delta Omega Honor Society for Public Health and awards named for former faculty members. The ceremony is held each year in conjunction with the James E. Grizzle Distinguished Alumni Award honoring a UNC Biostatistics alumnus who has made remarkable contributions within the first decade of their post-graduate career. Professor Grizzle was department chair from 1972 through 1987, noted for expanding the department through numerous faculty hires, establishing the Lipids Research Center (now known as the CSCC), and starting the BSPH program. The 2020 recipient of the Grizzle award was Dr. Yijuan Hu, 2011 PhD graduate and currently Associate Professor of Biostatistics at Emory University. The title of her talk was “Robust Testing for Differential Abundance in Microbiome Data.” Following Dr. Hu’s lecture, the following awardees were honored:

Campus View at Graduation Time

· Elaine Kearney Kowalewski – recipient of the Regina Elandt-Johnson award for Best Master’s paper. Elaine’s Master’s paper (2020), under the direction of Professor Gary Koch, was entitled, “Evaluation of a Method for Sample Size Re-Estimation for a Clinical Trial to Compare Two Test Treatments to Control.” Elaine is continuing a PhD student in the department.

· Sean McCabe – recipient of the Larry Kupper award for Best Dissertation Publication. Sean’s dissertation (2020), under the direction of Professors Mike Love and Danyu Lin, was entitled, “Statistical Methods for the Analysis of Multi-Omics and Multi-Study Datasets.” Sean is currently a post-doctoral fellow at Harvard University.

· Bonnie Shook-Sa – recipient of the Barry H. Margolin award for Excellence in Doctoral Research. Bonnie’s dissertation (2020), under the direction of Professor Michael Hudgens, was entitled, “Inverse Probability Weighting and Outcome Regression Approaches in Causal Inference and Survey Sampling.” Bonnie is currently an assistant professor in the department and working with the CFAR and CSCC.

The Bernard G. Greenberg Distinguished Lecture Series honors the first chair of Biostatistics (1949-1972) and former Dean of the School of Public Health (1972-1982). Each year, the Biostatistics faculty select a distinguished colleague to receive the Greenberg Distinguished Lecturer Award based on the quality and impact of their research (see past recipients of this award here). In May 20-21, we welcomed Prof. Xihong Lin of Harvard University as our Greenberg Lecturer. Professor Lin gave three lectures on her research in statistical and computational methods for massive data, spanning the application areas of whole-genome sequencing studies, genome-wide epigenetic studies, and Covid-19 epidemic dynamics. It was a pleasure to host Professor Lin for a virtual campus visit and lecture series this year.

Our 2021 graduates were able to attend an in-person commencement ceremony at Kenan Stadium on Friday, May 14. I was one of the many volunteers that day and loved recognizing some faces in the crowd. My video message to all of our graduates can be found here.

A heartfelt congratulations and a round of applause to all of our award recipients and new graduates!

For the second year in a row, the Joint Statistical Meetings (JSM) will be held virtually August 8 – 12, 2021. We are planning to host a UNC reception for faculty, students, and alumni on Monday, August 9, 2021, at 5:00 p.m. Please join if you can for what has always been a fun event in person, and a great opportunity to re-connect with colleagues. Details for joining the virtual platform are forthcoming. If you are a registered attendee, you can access event details in the JSM online program.

Enjoy the summer, and I look forward to seeing everyone back on campus in August.




2021 Phi Beta Kappa inductees announced

Students from the Gillings School’s Department of Biostatistics were recently inducted to Phi Beta Kappa, the country’s oldest and most honored college honorary society. Phi Beta Kappa was founded in 1776 as a society devoted to the pursuit of liberal education and intellectual fellowship. To this day, it continues to honor those who achieve excellence in a broad-based exploration of arts and sciences during students’ undergraduate years. Past and present Phi Beta Kappa members from across the country include 17 American presidents, 41 U.S. Supreme Court Justices, and many of our BSH’rs.

Phi Beta Kappa has 290 chapters nationwide. UNC Chapel-Hill’s chapter, Alpha of North Carolina, was founded in 1904 and is the oldest of the seven chapters within the state. Fewer than 1% of all college students qualify for initiation. Membership in Phi Beta Kappa lasts a lifetime, allowing employers to view it as a competitive advantage and a mark of personal distinction.

This semester, the Department of Biostatistics is proud to have four undergraduate students invited to join this prestigious honor society. The Spring 2021 inductees are Alexis Victoria Bell, Selen Pinar Gizlice, Mincen liu, and Natalie Poupart.



Click image to view University story

The University’s Biostatistics Department at the Gillings School has been at the forefront of Data Science even before the term was coined. We are excited that the University of North Carolina will continue this tradition with the new addition of a data science minor launching this fall.


  • August 18 – First day of classes
  • September 6 – Labor Day, no classes held
  • October 12 – University Day
  • October 21 and 22 – Fall Recess
  • More important dates can be found here.

UNC Chapel Hill Alumni Reception – All current students, faculty and alumni welcome to join.

Monday, August 9, at 5:00 PM Eastern

To join this event you must register in advance: https://uncsph.zoom.us/meeting/register/tJ0tf-Gvqj4sEtX6oIrypBopxoVk_YrqmGBW


If you have news or a story idea you feel would fit BiosBeat, please submit them to Jeff Oberhaus.

PLEASE NOTE: Given the recent events regarding COVID-19, the annual BiosRhythms is on hold. Issue 31 will not be sent until further notice.