ICompBio REU
Research Experience for Undergraduates in Interdisciplinary Computational Biology
The University of Tennessee at Chattanooga
The iCompBio REU in 2025 (funding pending) tentatively is from May 20 to July 28, 2025.
10 weeks in-person REU
(remote participation and flexible schedules can be accommodated)
Summer 2025 REU (funding pending) application is now open
The about 10-week iCompBio REU focuses on interdisciplinary computational biology training to undergraduates in science, technology, engineering, math, and other related fields. Please look through the potential research projects and then choose your preferences of the potential research topics.
- $6,250 stipend + housing and meal plan
- Travel allowance up to $500
- Inter-disciplinary training in computer science and biology
- Coding boot camp.
- Diverse range of research topics.
- Chattanooga is a great outdoor city
A one-week computing bootcamp will be provided to teach students essential data science using R, and advanced data processing and deep learning methods using Python. Each student will be jointly mentored by a computer science mentor and a biology mentor. The computing training includes coding, modeling, simulation, deep learning neural networks, computer vision, parallel computing, statistics, data visualization, and mobile App development.
Application Process
Student application includes an application form, transcripts, a personal statement, a resume, preferences of research topics, and at least 2 letters of recommendations. All application materials will be upload through an online-application form. Student applications will be reviewed by a team of faculty research mentors. (Please note that the application form is a Google-form which requires a Gmail account to complete).
Summer 2025 REU (funding pending) application is now open
Apply for the Summer 2025 ICompBio REU
Review of applications will start on March 15, 2025.
Please be aware the fund of the 2025 REU program is still pending, and we will update status of the 2025 iCompBio program as best as we can.
Contact Information
For more information about the program, please email: [email protected]
PI, Dr. Yingfeng Wang
Co-PI Dr. Jannatul Ferdoush
High-lights of Past REU Participants
Student publications
- Ledesma, Dakila, et al 2020, Application in Plant Sciences
- Baldwin, Q. and Panagiotou E., 2021, The local topological free energy of proteins (submitted) [link to preprint]
- Baldwin, Q., Sumpter, B. G. and Panagiotou E., 2021, The local topological free energy of the SARS-CoV-2 spike protein (submitted) [link to preprint]
Student presentations:
Student updates
- Zimmerman, Dell, Goldwater Scholarship, 2020
- Falvey, Cleo, NSF BIOREU blog, 2020
- Isimeme and de Blonk, NSF BIOREU blog, 2020
- Zimmerman, Dell, NSF BIOREU blog, 2020
Potential iCompBio REU Research Projects.
Using Autoencoder to Analyze Metabolite Structures (Yingfeng Wang)
Dr. Yingfeng Wang's lab is interested in applying machine learing tools to analyze bioinformatics data. Autoencoder has been widely applied to map the original data to another representation without information loss. This project will apply autoencoder to map the original metabolite structures to another data format for analysis. Students will develop an autoencoder by using PyTorch and conduct data analysis.
In Silico Functional Characterization of Hypothetical Proteins from Bacteria (Jannatul Ferdoush)
Hypothetical proteins (HPs) exist in bacteria which have unknown functions. In bacteria, about 50% of the genome contains HPs, which are implicated in disease pathogenesis. Therefore, it is important to identify and characterize HPs which would play important roles in understanding the mechanisms of disease, drug designing, antibiotic production etc. Hence, Dr. Ferdoush’s lab aims to find structure and functions of HPs using a number of bioinformatics tools. Students will employ several bioinformatics tools such as ProtParam, Cello, PSORTb, SOSUI, PSLpred, HHPred, Expasy, as well as tools from NCBI database to identify and characterize bacterial HPs for future potential therapeutic intervention.
Bio-Signal-Enhanced Machine Learning Control for a Hand Exoskeleton (Erkan Kaplanoglu)
In this project, students will investigate the development of a bio-signal-driven machine learning algorithm to control a hand exoskeleton. By leveraging bio-signals, such as electromyography (EMG), with advanced machine learning techniques, the system aims to enhance the intuitive control and functionality of hand exoskeletons for rehabilitation and assistance. This approach involves analyzing and interpreting bio-signals to adapt the algorithm in real-time to the user's intended movements, resulting in more natural and responsive interactions with the exoskeleton. Findings from this research will provide insights into the potential of bio-signal-driven machine learning algorithms to optimize assistive device control, improving their effectiveness and usability across various applications.
Bio-Inspired Design and Control of a Multi-Finger Robot Gripper (Gokhan Erdemir)
Why are certain robotic grippers more efficient and adaptable than others, even with similar structural designs? How can we enhance the versatility of robotic grasping mechanisms? Dr. Erdemir’s lab and his team investigate these questions by applying bio-inspired design principles and developing advanced control algorithms. They explore the mechanics of multi-finger grippers based on animal structures through mathematical modeling and computational simulations, generating predictive models to improve grasping capabilities. Dr. Erdemir’s group also develops machine-learning algorithms to optimize robotic control for versatile applications. Students in this project will learn network modeling, control algorithm development, and machine learning techniques for robotics.
Molecular dynamics studies on bacterial cell membranes and associated proteins. (B. Harris)
The membranes surrounding a bacterial cell are responsible for mediating processes such as cellular recognition, signal transduction, and the transportation of ions and molecules across the membrane. Further research into these processes is essential in order to advance our understanding of how pathogenic bacteria sense and adapt to their environment. Students working in Dr. Harris’s lab will conduct molecular dynamics studies of bacterial cell membranes and analyze how membrane phospholipid remodeling contributes to survival and persistence. This research is essential in order to advance strategies for the prevention and treatment of diseases caused by bacterial pathogens.”
Use remote sensing and GIS to study the impact of urbanization on surface water quality (A. Hossain)
The aim of this research is to use remote sensing and GIS to determine whether increased urban development in Hamilton County has had a negative effect on surrounding water quality and if this interaction is a concern for urban sustainability. Students will learn to perform various machine learning methods to analyze geo-location, environment al and weather related data.
Automatic assessment of soil-borne fungal growth rates using time-lapse images ( D. Beasley)
A critical goal of modern ecology is to determine how rapid climate change will influence ecological interactions among species. Of these, fungal interactions are of particular importance as major drivers of population dynamics, community structure, and biological diversity. Here, we propose to work with undergraduate students to explore the effects of temperature on soil-borne fungi by quantifying growth rates using high resolution time-lapse images. In addition to understanding implications for insect-fungal interactions under changing ecological conditions, the project will provide students with computational research training.
Mathematical and computational modeling for infectious diseases (Jin Wang)
Infectious diseases remain a serious public health burden throughout the world, leading to high morbidity and mortality every year. Mathematical modeling provides a theoretical tool to investigate the mechanisms of infectious diseases and offers useful guidelines for the design of control strategies. In this project, REU students will learn various techniques in constructing mathematical models, analyzing disease dynamics, and numerically simulating disease transmission and spread. In addition, the developed mathematical and computational epidemic models will be possibly applied to realistic case studies.
Apply graph algorithms to ecological big data to identify key factors that influence mating and parental dynamics (H. Klug)
Students will use network control theory to develop novel computational tools that allow us to effectively utilize large datasets to answer complex, broad, and fundamental questions in ecology and evolution. Students will apply these tools to ask a fundamental question in evolutionary ecology: How do life-history traits, ecological conditions, and sociality interact to influence mating and parental dynamics?
Rare versus common plant species (J. Boyd)
The question of why some species are rare while others are common is enduring and has important implications for ecological theory, rare species conservation, and overall biodiversity. As part of a broad research agenda to help address this question, I am using network analysis – a computational tool that uses mathematical graph theory to link different concepts as a visual map – to characterize the body of research comparing rare and common plant species, an area that has not been comprehensively reviewed in nearly two decades. One of the aims of this work is to refine techniques to facilitate the use of network analysis to characterize broader areas of scientific research. REU students will learn and apply network analysis to explore an area of scientific research of their interest.
Deep Learning Techniques for Recognition and Anomaly Detection of Human Gaits (Ziwei Ma)
Human gaits can be used to evaluate individual’s overall health condition. Wearable sensors are able to capture motion with more precise and detailed information at lower cost than traditional clinical measure procedures. How to decode wearable sensors signals? How to provide efficient and accurate motion detection using these sensor data? Dr. Ma explores the human gaits recognition and anomaly detection using statistical models and deep learning algorithms to extract information from sensor data. Students will learn statistical modelling and deep learning algorithms for sensor data.
Mobile app and deep learning for plant species digitization (J. Shaw)
Large digital biology data sets are important for increasingly trans-disciplinary studies Dr. Shaw’s group led a recent large digitization effort of herbarium collections, aiming to digitize 886,373 vascular plant specimens housed in 12 separate herbaria in Tennessee. Students will participate in developing a field data collection program, designed by digitization staff to enforce input normalization. Students will learn to apply recurrent neural networks using the Python Keras library to detect sample classification anomalies using the large number of plan digital images.
Evolutionary plasticity of fatty acid pathways in microbial genomes (D. Giles)
The Giles laboratory has observed the effects of fatty acids in several bacteria of medical importance, including Vibriospecies, Pseudomonas aeruginosa, Acinetobacter baumannii, and Klebsiella pneumoniae [24]. Our central hypothesis is that the conservation of fatty acid handling machinery stems from an aquatic origin. This project will address this hypothesis bioinformatically by phylogenetically mapping Gammaproteobacteria based on possession of genes encoding fatty acid transporters (FadL), fatty acid CoA synthetases (FadD), and acyltransferases (PlsB/C/X/Y). Characterizing the extent to which groups of this class of ecologically and medically important bacteria can import, activate, and assimilate exogenous fatty acids may reveal evolutionary insights.
Modeling, Monitoring, and Mapping of Urban Infrastructure (D Wu, Y Liang, and L Yang)
Real-time and automated modeling, monitoring and mapping of urban infrastructure, such as, roads, bridges, and subsurface utilities, are essential for maintainability and serviceability of these important assets. Dr. Wu’s team is interested in creating a digital twin of urban infrastructure by combining geophysical instruments, such as, ground penetrating radar (GPR), light detection and ranging (LiDAR), and hyperspectral imaging sensors, with computing technology, such as, artificial intelligence, augmented reality, and cognitive decision-making. The digital twin, considered as a digital replica of physical infrastructure, will be used to simulate, monitor, and optimize the operation and maintenance of those physical infrastructure assets. Students will learn the knowledge, skills, and tools related with cyber-enabled high-bandwidth sensing and information processing.
Molecular dynamic studies on lysozyme modifications (J. Kim)
Dr. Kim’s laboratory is interested in benzoquinones (BQs) and their biological roles because BQs are available ubiquitously in the environment as free quinones, protein cofactors, or as metabolites of polycyclic aromatic hydrocarbons. BQs exhibit toxicity by reacting with cellular proteins and nucleic acids mainly through redox cycling. Recently, our lab discovered BQs can induce protein aggregation which is known to associate with neurodegenerative diseases. We will investigate how BQs interact with a model protein, using a molecular dynamics approach.
Cancer genomics (L. Gao)
Cancers are distributed unevenly across the body, but the importance of cell intrinsic factors such as stem cell function in determining organ cancer risk is unknown. Prom1+ Cell (a commonly used cancer stem cell marker) generative capacity is a major determinant of organ cancer risk. To understand what cell properties might dictate cancer risk, we use statistical modeling to explore how Prom1+ cell population size, proliferative capacity, and generative capacity in each organ related to susceptibility to tumorigenesis.
This figure shows that generalized linear mixed models with a logit link (GLMM_LL) of tumor probability versus Prom1 cell generative capacity, prom1 cell population size and prom cell proliferative capacity (A-C). Numbers at top of each graph indicate the regression coefficient and p value of each as well as the AIC score (D and E) GLMM-LL iterative multivariable modeling of tumor probability versus Prom1 cell generative and proliferative capacities in adult and neonatal tissues.
Biodiversity estimates of microscopic organisms in the Tennessee waters (F. Leasi)
This research aims to investigate biodiversity shifting of microscopic organisms (invertebrates, bacteria, fungi, protists, etc.) in changing environments, with a focus on freshwater. Freshwater ecosystems are under intense pressure from overuse, pollution, and habitat degradation. Climate change is predicted to add additional threats and interact in complex ways with other stressor types, such as eutrophication. Consequentially, the distribution of sensitive freshwater species is expected to change. This is especially true for microscopic organisms, which are abundant and ubiquitous in aquatic ecosystems, performing key functions such as nutrient cycling and sediment stability. Yet, their unexplored diversity and response to disturbances represent one of the major challenges in biology and currently limits our capacity to understand, mitigate, and remediate the consequences of pollution and environmental change. In this project, students will integrate field investigation, genetics, genomics, and computational approaches to test the following hypotheses: 1- biodiversity of microscopic organisms is correlated to environmental conditions; 2- biodiversity shifting can be predicted in changing environments; and 3- the interaction invertebrates and microbial communities significantly affect biodiversity shifting.