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!
Highlights of this SPECIAL DOUBLE Issue:
Conference Celebration of Bios’ 70 Years
Professor Annie Green Howard Profile
BSPH Summer Fun Stories
Gillings Graduate Research – Research on Epigenomic Marks
Department’s Research Published Recently in Lancet, Biometrika and More
Huge Award to Join Fight Against National Opioid Crisis
Conference Celebration of Bios’ 70 Years
The last week of October, a special two-day reunion merged the work of current students and alumni. The festivities started with a well-attended student poster session in the Armfield Atrium in the Michael Hooker Research Center. Over a dozen biostatistics students (undergraduate and graduate) captivated the audience of students, faculty, staff, and alumni with their research work.
The poster session included the inaugural presentation of the Koch Bios poster award, created to honor Professor Gary Koch for his commitment and dedication to our department’s students. Two awardees were announced, graduate students Jesus Vazquez and Sean McCabe. Vazquez presented a poster on Validity of the Global Physical Activity Questionnaire (GPAQ) in Hispanic/Latinos in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL), and McCabe presented on A Latent Dirichlet Model to Compare Expressed Isoform Proportions to a Reference Panel.
The conference continued at the Carolina Club at UNC George Watts Hill Alumni Center after the morning poster session. Department Chair Michael Kosorok welcomed all and introduced keynote speaker, Dean F. DuBois Bowman (PhD 2000). The conference sessions included topics such as machine learning, causal inference, precision health, and non-parametrics, to highlight a few. Sessions, which included no faculty, were presented by current students and alumni. They emphasized the interconnectedness between the work alumni have done and the work that current students are in the process of completing. These inspiring presentations showed the foundation that the Gillings School’s biostatistics department provides for its students and their future endeavors.
The evening banquet featured keynote speaker Dennis Gillings, PhD, CBE, who provided an insightful view into the rich history of the department with photos and other visuals from its earlier years. Gillings detailed how far the department has come since its founding in 1949. In the 1951-1952 school year, faculty members for the entire School of Public Health fit on just one page. Today, the biostatistics department has 41 faculty members on its own. At the department’s founding, the emphasis of biostatistics was on vital statistics and public health statistics. Today, we see a growing emphasis on big data, as well as applications of biostatistics across life and environmental sciences. As Gillings mentioned, the field of biostatistics, as well as our department, is continuing to grow and will provide even more opportunities in the future.
This two-day celebration was a wonderful way to reminisce what the department has done, and all that it continues to do. Two outstanding recent alumni, Harvard Professor Briana Stephenson (PhD 2019) and current graduate student Katherine Gora Combs (BSPH 2019) closed the event with a reminder that those in the audience would need to lead the next conference reunion and give back to the Gillings School.
SPOTLIGHT ON FACULTY
Annie Green Howard, PhD UNC Chapel Hill 2012
Professor Position: Associate Professor of Biostatistics
Time at the Gillings School: 7 years as faculty with 7 before as a graduate student
What I do (and why I love it): My work is primarily based out of the Carolina Population Center, working to help better understanding the complex pathways to cardiovascular disease. The specific focus of my research has involved the incorporation of genetic, -omics, environmental, and lifestyle behaviors into an epidemiologic and biostatistics framework, longitudinal analysis methods in observational cohorts experiencing dramatic environmental changes and novel methods to characterize CVD risk factors, specifically physical activity and weight. I feel an important aspect of being a successful researcher and one of the things I love most about my job is working on interdisciplinary research teams with people who have a wide range of statistical knowledge and experience. Not only do I learn so much from them, but it makes for a fun work environment to have people who have similar passions and are as excited to learn from me as I am from them.
First job or internship I had was: Worked with the Crash Injury Research and Engineering Network (CIREN), a group which seeks to determine injury causation in car crashes, as an administrative assistant to help gather and organize data as well as prepare presentations of the data for the team’s review. This included going with team members to interview patients, take photographs, and collect details from both the car and crash site.
GET TO KNOW BIOS
What to do in the Summer? Exotic Island Water Adventure or Crunch Numbers in Washington
Water Study in the Galápagos – Gillings School students and staff connected with students from Universidad San Francisco de Quito where they conducted a study in the Galápagos Islands of Ecuador. Students with backgrounds ranging from nutrition and environmental health to anthropology and biostatistics collected various types of data from households across the island. In addition to the times, they relaxed on the beach, hiked scenic trails, and mountain biked down a volcano, the students analyzed tap water, implemented household surveys, and gathered anthropometry measurements to assess health outcomes.
Donald Fejfar, a UNC Biostatistics BSPH’r, explains how he applied his biostatistics skills this summer, “Although my knowledge in biostatistics is still young, some of the really fundamental things that you learn early on, were the biggest bits of help. Knowing how to manage data and code for statistical programming helped me manage the data files. Having a background in making analysis plans before fieldwork and grant writing begins helped us get a head start on what we needed to collect and how we were going to use it. Biostatistics also just helped me in designing a study from the get-go. None of these are incredibly complicated or high-level biostats topics, but they were crucial to our project.”
The main objective of this study was to determine how to provide clean water to residents. Biostatistics proved to be an integral part of achieving that goal. Besides, students gained hands-on experience that speaks to the essential role public health knowledge plays in any future plans the students may have.
Intern in the Nation’s Capital – Over the past summer undergraduate student, Jane Williford had the amazing opportunity to be a Junior Fellow through the University of Maryland and University of Michigan’s Joint Program for Survey Methodology. Jane lived in DC and worked full time at the National Center for Health Statistics (NCHS) within the Division of Research and Methodology (DRM), while also working with staff from the National Health Interview Survey (NHIS). Jane was exposed to how the government’s statistical agencies work as well as cutting edge research relating to survey methods and biostatistics at NCHS. She also visited other government and non-government statistical and survey agencies in the Washington Metro Area.
Jane highlighted how one of the best parts of her internship was being able to work closely with her supervisor, Dr. Jennifer Parker, as well as many other talented individuals at NCHS on different projects of interest. She aided in conducting hands-on research by doing things like comparing different forms of data collection and working with people from the Division of Health Interview Statistics. She also came up with an idea for an MMWR Quickstat, and was encouraged to create the Quickstat from start to finish using NHIS data. This work was published last month by the Centers for Disease Control (CDC).
Throughout the summer Jane improved her SAS programming and statistical skills, her data and communication skills, and her overall professional confidence. She also gained an understanding of an area of research she had never been exposed to. Jane entered the Gillings School’s Biostatistics BSPH program wanting to use her quantitative abilities to try and tackle some of the world’s biggest health problems. Today, she is more driven and excited than ever to pursue stimulating and meaningful work in this area of public health.
To read Jane’s QuickStat published by the Center for Disease Control and Prevention, click here.
Chromatin immunoprecipitation followed by next-generation sequencing (ChIP-seq) is a technique to detect genome-wide regions of protein-DNA interaction such as transcription factor (TF) binding sites or regions containing histone modifications. This impacts gene expression and influences biological processes. ChIP-seq experiments can be used to understand epigenomic mechanisms in which TFs and histone modifications play an important role. This is used to explain heterogeneity at both the molecular level (gene expression, gene silencing, etc.) and on an individual level (cancer incidence, cardiovascular disease, etc.). In cancer research, histone modifications have been shown to play an important role in carcinogenesis, progression, and tumor suppression.
This study was conducted by Pedro Baldoni, a doctoral student at the Gillings School. Baldoni has received the University Cancer Research Fund (UCRF) School of Public Health Student Award for 2019. He is working in the Biostatistics Department’s Collaborative Studies Coordinating Center (CSCC) on the Hispanic Community Health Study / Study of Latinos (HCHS/SOL) project in addition to pursuing his degree. The work he conducts on the project builds on his academic interests and his experience as a research assistant allows him to see many aspects, including allowing him to lead manuscripts.
ChIP-seq experiments help identify genomic locations of origin, these genomic coordinates contain a high density of mapped reads, referred to as enriched regions, where a protein of interest was bound to DNA. These are then identified through statistical analysis. All other genomic positions pertaining to non-enriched regions are referred to as background. However, the detection of enrichment regions in ChIP-seq experiments is not easy for reasons like the diversity of enrichment profiles, serial correlation in the data, and sample-specific characteristics.
To approach these challenges, researchers presented a Zero-Inflated Mixed-effects Hidden Markov Model (ZIMHMM) to analyze data. The ZIMHMM is tailored to call broad peaks in consensus across multiple ChIP-seq technical or biological replicates. When used by researchers, it modeled the excess of zeros and accounts for sample differences. Methods focusing on peak calling from multiple samples are growing of interest since there has been a reduction of DNA sequencing costs and higher data availability. The results from using the ZIMHMM could trigger new insights to investigators interested in detecting something like cell-specific activated genes.
Published in Biometrics read the full article here.
There has been a lack of evidence-based research when developing treatment regimens in children who are newly diagnosed with ulcerative colitis (UC), a chronic form of inflammatory bowel disease (IBD). The PROTECT study aimed to test the safety and effectiveness of the two standard treatments of UC- mesalazine and corticosteroids.
PROTECT is a multi-center study, the Collaborative Studies Coordinating Center at the Gillings School of Global Public Health is the study’s data and Statistical coordinating center. Sonia Davis, DrPH, praised the Department of Biostatistics’ contribution, “The CSCC was vital to the success of this complex precision health study, co-authors include our Nathan Gotman, Alison Marquis and recent Biostatistics graduate students, Bradley Saul and Jessie Wang.”
Over five years, PROTECT sought out pediatric patients ages 4-17 across the US and Canada that were newly diagnosed with UC. To better understand the effects of genetics, mechanisms of inflammation, Vitamin D, and bacteria, bio specimens such as blood, stool, and colonic biopsy tissue were obtained. Patients then received either mesalazine or steroids for initial treatment, with pre-established protocols in place for escalation to more advanced therapies like thiopurines or anti-tumor necrosis factor-α (TNFα) therapy. The primary outcome was steroid-free remission at 52 weeks. The secondary outcome was an advancement to immunomodulators (i.e. the advanced therapies like anti-TNFα therapy) or colectomy.
Researchers found that those at low baseline clinical severity with clinical remission at week 4 of treatment were significantly associated with 52 weeks of steroid-free remission. This is significant in deciding whether certain treatment alternatives should be considered for specific UC populations. They also found that patients with low baseline levels of Vitamin D were more likely to escalate to more advanced therapies such as anti-TNFα therapy. Shifts in the microbial community and intestinal gene expression have also been linked in predicting the likelihood of response to biological therapies. Decreased expression of genes and increased expression of antimicrobial peptides were linked to an increased likelihood of escalation to anti-TNFα therapy. The findings of this study will help in developing more personalized patient approaches when treating UC.
Read the full article here.
Students Show Off Research at 70th Celebration Conference
Gillings School researchers join fight against national opioid crisis
The Gillings School has recently been selected to lead one of the 375 grant awards that make up the NIH Helping to End Addiction Long-term (HEAL) Initiative. This research aims to reduce the rate of opioid overdose, as well as improve treatments and achieve long-term recovery from opioid addiction.
At the frontline of this $51.781 million “Back Pain Consortium (BACPAC) Research Program Data Integration, Algorithm Development and Operations Management Center (DAC)” research project is a professor and associate chair in the Gillings School’s Department of Biostatistics Lisa LaVange, she is also the Director of the CSCC. The BACPAC research program focuses on translational, patient-centered efforts to address the need for effective and personalized therapies for chronic low back pain, one of the most common forms of chronic pain among adults worldwide. In 2016 alone, 50 million U.S. adults suffered from chronic back pain. Since current treatments are ineffective, this has contributed to the increased use of opioids. In 2018, an estimated 10.3 million people 12 years and older in the U.S. misused opioids, including heroin.
Anastasia Ivanova, PhD, professor of biostatistics at the Gillings School, is the co-principal investigator for the project, which will run through May 31, 2024.
As LaVange expresses her excitement on receiving this grant she adds, “As part of this important initiative, we will work with mechanistic research centers, technology research sites, and phase 2 clinical trial groups to deliver an integrated model of chronic lower back pain and explore innovative technologies. Our team includes experts from the Gillings School, the UNC School of Medicine and UNC’s Lineberger Comprehensive Cancer Center, and the investigators represent several different academic departments across campus.”
NIH Director Francis S. Collins, MD, PhD, who launched the initiative in early 2018 comments, “It’s clear that a multi-pronged scientific approach is needed to reduce the risks of opioids, accelerate development of effective non-opioid therapies for pain, and provide more flexible and effective options for treating addiction to opioids. This unprecedented investment in the NIH HEAL Initiative demonstrates the commitment to reversing this devastating crisis.”
See the full post on the Gillings School website here.
AWARDS / HONORS / FACULTY NEWS
Doubly truncated survival data arise when a failure time T is observed only if it lies within a time interval. The nonparametric maximum likelihood estimator for doubly truncated data was introduced by Turnbull (1976) and is widely used in areas such as astronomy. The nonparametric maximum likelihood estimator is widely used to estimate the underlying failure time distribution. Using a directed graph representation of the data, a certain graphical condition holds if and only if the nonparametric maximum likelihood estimate exists and is unique. If this condition does not hold, then such an estimate may exist but not be unique, so another graphical condition is proposed to check whether such an estimate exists.
This study is co-led by Michael Hudgens a biostatistics professor at the Gillings School and is the director of the Biostatistics Core of the UNC Center for AIDS Research (CFAR). He has experience in collaborative research and statistical methodology development related to studies of infectious diseases. He has co-authored more than 200 peer-reviewed papers in numerous statistical journals. He is an elected fellow of the American Statistical Association and has taught graduate level biostatistics courses at UNC for over ten years.
Before July 1, 1968, a study was conducted by the United States Centers for Disease Control and Prevention of individuals infected with HIV through blood transfusion and diagnosed with AIDS. The dataset considered for this study contained 34 children that were ages 0-4 at the time of the blood transfusion. A study recently analyzed these data to estimate the distribution of age at onset of AIDS, which is assumed to be doubly truncated by age at contaminated blood transfusion. The data was first observed through the Efron–Petrosian iterative algorithm and then a directed graph, which was constructed using the Rpackage igraph. The is.connected function from igraph can then be used to determine that the graph is connected, but not strongly connected.
In either case, because the nonparametric maximum likelihood estimate does not exist for these data, both the Efron–Petrosian estimate and the modified estimate can potentially be misleading. Constructing the second graph was critical in assessing whether it is strongly connected, connected, or not connected. It is important to avoid misleading estimates when computing the nonparametric maximum likelihood estimator in the presence of double truncation. The proposition indicates that misleading estimates will only tend to occur in smaller samples.
Read the full article here.
Gillings Students Receive Fulbright Awards – BSPH Graduate Gancz
Three Gillings School students have received the Fulbright United States Program Awards. Among them are Melissa Stockton, Erin Danford, and Biostatistics BSPH graduate (2019) Abigail Gancz. The three UNC Gillings scholars will join 20 other UNC-Chapel Hill students and graduates, and more than 2100 U.S. citizens, who will study, conduct research, or teach abroad during the 2019-2020 academic year through the Fulbright Program.
Read the full article here.
Bios 70th Look Back
Greenberg Report 1967
In the mid-1960’s Professor Greenberg chaired a committee appointed by the National Institute of Health (NIH) National Heart Advisory Council and directed to formulate a process for large, multi-centered clinical studies. The result is now known as the “Greenberg Report.” The Report is about the organization and administration of cooperative studies, highlighting the importance of positions such as the Chairman, Policy Advisory Board, Executive Steering Committee, a Coordinating Center, and data-contributing participants. The need to communicate effectively with the Coordinating Center to conduct a cooperative study and control of performance at all levels is emphasized.
The Gilling School’s Epidemiology Professor Wayne Rosamond describes the Greenberg Report: “The thoughts expressed in the Greenberg Report more than 50 years ago are still relevant to today’s collaborative studies in the cardiovascular field. The comments from the report that resonates most to me include the guidance that collaborative studies need to have a sharp focus on the ability to translate research progress into clinical and preventive practice. The high value this document places on continued strong leadership within a well-defined organizational structure is clear as its vision that a coordinating center is the focal point of this essential element for success.”
The Greenberg Report created the foundation for numerous studies that have been conducted since its publication. It is still used today as a guideline for various collaborative studies, and many rely on its content for understanding the standard. Upon creating this report, Bernard Greenberg was paving the way for many researchers. The influence of this report extends far beyond our campus.
Important Student Date Reminders
Dean’s Lecture Series: Human Rights in Global Health November 14 @ 3:00 pm – 4:00 pm LINK
The purpose of this mini-symposium/networking event is to advance collaborations between the Schools of Medicine, Pharmacy and Public Health by encouraging new and cross-discipline research endeavors. Register here.
If you have news or a story idea you feel would fit BiosBeat, please submit them to Jeff Oberhaus.