iMPACT

Two new Research Experiences for Undergraduates (REU) in the field of statistics are training students in cutting-edge and advanced data analytics and computational skills essential to interdisciplinary research across the fields of statistics, microbiology and quantitative sciences. 

New grant trains students in data analytics

The Department of Statistics was awarded its first Research Experiences for Undergraduates (REU) grant this year. The National Science Foundation’s REU program supports comprehensive, hands-on research experience for undergraduate students in the STEM fields, and awards funds to initiate and conduct projects that engage a number of students in research.

Statistics faculty Yanming Di, Lan Xue, Thomas Sharpton (PI; joint appointment in microbiology), Duo Jiang and Yuan Jiang received the NSF REU grant from the American Statistical Association (ASA). The $380K project is for 2016-2018, to support three REU sites per year, for a total of nine REU sites across the country. The grant funds 10 weeks of research and training for four undergraduate students at each site. Each student receives a stipend of $8,000.

The overall objective of the ASA-supported REU program is to promote undergraduate research experiences in statistics and to prepare students for graduate study in statistics. According to ASA, “The students will see how statistics has an impact on fields such as engineering, atmospheric science, health care, and all kinds of public policy.”

Statistics REU students (from l to r) Aaron Huang, Ellen Kulinksy, Betsy Hensel and Shelby Taylor enjoy the outdoors in Corvallis.

Oregon State University was chosen by ASA as one of three REU sites this year and is currently hosting students from across the country from June 19 until August 25. The four REU participants were selected from more than a hundred applicants in a highly competitive process.

The REU students are Shelby Taylor from Brigham Young University, Aaron Huang from the University of Washington, Ellen Kulinsky from the University of California-Berkeley and Elizabeth Hensel from the University of Virginia.

“The students are stellar and they are doing a fantastic job with the research. They were selected for this transdisciplinary REU because of their prior knowledge and preparation in both statistics and biology. We are very lucky to have them here,” said Sharpton.

The students are gaining exposure to the entire data analysis process as it relates to biological research. They analyze DNA sequence data and use statistical methods to determine how the types of bacteria that live in the gut, known as the gut microbiome, relate to human health, lifestyle and environmental conditions.

Specifically, the students are analyzing data from a large, crowd-sourced, citizen science project led by American Gut which collects human samples, ranging from saliva to stool, along with questionnaire responses about individual lifestyles and diet. Working closely with an interdisciplinary team of faculty, each student will conduct a complete data analysis, which includes data quality control, applying statistical and bioinformatics techniques, data visualization. Faculty members will train students in all of the techniques and skills that they need to complete the project.

Students will analyze the data to gather information on how the gut microbiome varies across individuals and its association with a variety of health and lifestyle factors. Some of the REU projects explore how body mass index affects the microbiome; the relationship between age and the microbiome; and the associations between the gut microbiome and food and alcohol consumption.

REU students acquire knowledge and expertise in statistics and biology through intensive subject lectures. Hands-on experiential learning projects in bioinformatics and biostatistics give them ample opportunities to apply their theoretical and conceptual learning to design experiments and deduce results from complex data sets. The data-driven REU will prepare students to capitalize on the growing professional opportunities in data analytics.

While Sharpton and Jiang are extensively involved in guiding student research projects, the REU is a deeply collaborative process in which the other faculty are serving in important mentorship roles and providing expertise in theoretical and computational subject areas. Di, for example, has been teaching students computer coding and programming. Students are also learning R, LaTeX and Matlab as a part of their statistical and biological research.

“The REU has provided our faculty an opportunity to work with ambitious and talented undergraduate students. It has also exposed students to cutting-edge microbiome science. Through their research they are seeking answers to novel questions on human health and the microbiome, a new area of study about which much remains to be done,” said Sharpton. He hopes to bring a similar opportunity to OSU students in the future.

This REU in the area of microbiome informatics research is a pivotal part of Oregon State Microbiome Initiative (OMBI), and is slated to advance education and research in the statistical, biological and computational sciences at OSU. OMBI launched this spring.

The students’ research and learning experiences for the REU project are generously supported by the Department of Statistics, the College of Science and the Center for Genome Research and Biocomputing (CGRB). At CGRB, the students have conducted bioinformatics research and analysis with the aid of its biocomputing cyber infrastructure.

More information is available online.

Photo above: College of Science Dean Sastry Pantula with (l to r) Statistics REU students Ellen Kulinsky, Aaron Huang, Betsy Hensel and Shelby Taylor

The Summer Institute of Statistics targets talented, underrepresented students

Professor Javier Rojo  (far right), who joined the Department of Statistics in January 2017, has moved his award-winning REU site, Research for Undergraduates Summer Institute of Statistics to Oregon State (RUSIS@OSU) from the University of Nevada, Reno. In 2003 as a professor at Rice University, Rojo started the country’s first Research Experiences for Undergraduate (REU) Program in the field of statistics, which has been extremely successful in recruiting, training and guiding underrepresented minority and economically disadvantaged students towards advanced degrees in mathematics and statistics.

The Institute conducts a 10-week intensive summer program for the study of statistics and its applications for a cohort of 12-15 students every year. This summer 12 students, chosen from a pool of 70 applicants, are working on statistical research projects at OSU from June 19-August 24. The REU cohort is 50 percent female and 10 of the 12 students are underrepresented minorities. They hail from institutions such as the University of Texas, El Paso, University of Arizona, Occidental College, California State University, Bakersfield, Fresno State University, Texas State University, San Marcos, Duke, and Harvard among other places.

“One of the benefits of transferring the REU program to OSU is the name recognition that will attract talented students. It is also a great recruitment tool and may inspire students to apply to the Statistics Graduate Programs at OSU,” said Rojo.

There is promising data that REUs have a positive impact on graduate recruitment in host institutions. According to Rojo, approximately 10 students who received their Ph.D.s at Rice University had been RUSIS participants during the years the program was housed at Rice.

Current RUSIS students represent a mix of majors including engineering, computer science, mathematics and the social sciences such as psychology. Under Rojo’s guidance they are pursuing research on ambitious and exciting projects that involve studying data to measure the impact of the Clean Air Act on environmental pollution, investigating studies on the impact of obesity on the environment from a statistical standpoint, using probabilistic and statistical components to model data for better financial investment decisions as well as various other research projects.

Research projects at RUSIS, Rojo points out, involve heavy computation. The students are undergoing valuable computational training and learning various programming languages such as R, LaTex, Matlab and Mathematica taught by graduate students in the Department of Mathematics. Students also go through a four-week course on statistics and probability that brings them up to speed with statistics.

At the end of the program the students are expected to produce a technical report in LaTex and present a research talk to a scientific advisory committee comprising experts from Georgia Institute of Technology, University of Arizona, University of Michigan, Rice University and the University of Texas, El Paso.

Owing to the paucity of statistics undergraduate programs in the country, Rojo has encountered fewer than one percent statistics majors at RUSIS. Most RUSIS participants come from fields such as biology, business and computer science. But he notes that nearly 30 RUSIS alumni have gone on to earn a Ph.D. in statistics and biostatistics.

“My main objective is to encourage students to obtain a Ph.D. in statistics if they have the opportunity to do so,” said Rojo.

The program has been variously supported and funded by the NSF and the National Security Agency (NSA) for the last 15 years. Owing to Rojo’s sustained efforts and leadership, the American Mathematical Society (AMS) selected his REU program for its award “Mathematics Programs That Make a Difference” in 2014.

The AMS award citation states, “As the first REU in Statistics, RUSIS has served as a model program for others to emulate, both by encouraging undergraduates to pursue graduate studies in the mathematical sciences and by increasing the numbers of underrepresented minorities and women in mathematics and statistics.”

Under Rojo’s leadership, the program has taken phenomenal strides: After 10 years, the REU program reported that 85% of the undergraduates who attended the Summer Institute were admitted to Ph.D. programs around the country, with roughly 61% of students hailing from underrepresented populations and 53% of the participants have been female.

These impressive results were achieved through “intensive statistics courses, close supervision of research projects and visits to various research institutes and agencies in the area” according to Rojo, who is responsible for the students’ computational training and research projects.

Read more: Internationally renowned statistician joins faculty