Michigan Tech has received $4.3 million as part of an $18.5 million, two-year federal grant project to collect and analyze genomic data to address emerging infectious disease threats and enhance the state’s ability to respond to those threats. The funding will increase sequencing capacity in the state, starting with SARS-CoV-2, then expand to other infectious disease threats with potential for broad community impact. Two College of Computing faculty members are involved in the project.
Robust and scalable computational infrastructure and biometrics workflows.
Dr. Dukka KC, College of Computing associate dean of research and associate professor in the Department of Computer Science, is a co-Principal Investigator on the project. The aims of his work on the project are twofold: developing robust computational infrastructure and a bioinformatics workflow/pipeline for genomic surveillance.
Dukka will develop a robust and scalable computational infrastructure to build bioinformatics capabilities in Michigan, which is intended to improve timely genomic surveillance of infectious disease threats and alert the public of detrimental SARS-CoV-2 variants in Michigan. The work will involve developing an appropriate computational infrastructure for supporting this important task.
“Computational infrastructure can be a bottleneck in genomic surveillance. This hardware will be configured such that these types of large-scale bioinformatics workflows are well supported,” he says.
Dukka will also work to develop pathogen-specific bioinformatics workflows and pipelines, one of the most important components of genomic surveillance. He explains that these workflows and pipelines will define best practices to process raw read data, assemble viral or bacterial genomes, and construct read and genome alignments for downstream bioinformatics analyses.
“To accomplish these tasks, we will develop software pipelines to accelerate the sequencing read data preprocessing, de-novo genome assembly, and read-to-genome alignments and bioinformatics and mathematic analyses,” says Dukka. “As the size of the genome of these pathogens and their mutation rate varies, the workflows and pipelines for these different types of pathogens are deemed to be different.”
The immediate priority for this pilot project is to develop SARS-CoV-2-specific pipelines, then initialize pipelines for other infectious diseases. In the long term, researchers will work to develop additional pathogen-specific (SARS-CoV-2/bacteria-/fungi-specific) pipelines.
A disease population and outbreak surveillance infrastructure model.
Dr. Guy Hembroff, associate professor in the Applied Computing department and director of the Health Informatics graduate program, is a Senior Personnel on the new project. Hembroff has expertise in developing and managing AI prediction models for large-scale medical
data and images.
Hembroff will work to establish a robust, secure, and responsive outbreak surveillance infrastructure model for the public health disease population in the rural western Upper Peninsula, which comprises the counties of Keweenaw, Houghton, Ontonagon, Gogebic, and Baraga. The model will facilitate a partnership among Michigan Tech, healthcare providers, and federal and state agencies to safely transport and efficiently report environmental, human, and animal information and samples within the Upper Peninsula region.
In the infrastructure model, epidemiological data including reportable labs, ADT messages, and pathogen genomic data will be integrated and analyzed together in a timely manner. Michigan Tech will build a secure pipeline to receive collected regional data in a de-identified form—in near real-time—from health information exchanges.
Hembroff says that machine and deep learning models will be developed to improve regional health surveillance information. The infrastructure is expected to result in, 1) an improved reporting of epidemiology and pathogen surveillance that reinforces community engagement; 2) an improved picture of the health status of residents, which can learn to enhance genomic surveillance, data-driven decisions concerning resource allocation, needed services, and intervention priorities; and 3) provide evidence of the model’s success to properly evaluate its potential use in other areas of the state of Michigan.
Others on Michigan Tech’s research team include Principal Investigator Caryn Heldt, director of the University’s Health Research Institute; co-Principal Investigator Hairong Wei; and Senior Personnel Kristin Brzeski, Kelly Kamm, Stephen Techtmann, and Kui Zhang.
About the College of Computing Researchers
Dr. Dukka KC’s research is focused on various aspects of bioinformatics, including developing computational tools to decipher protein sequence, structure, function, evolution relationship; developing deep learning-based approaches for post-translational modification site prediction; GPU-parallelization of tools in bioinformatics; and development of tools and pipelines for Next-generation sequence analysis.
Dr. Guy Hembroff’s research is focused on machine learning, deep learning, intelligent medical devices, and cybersecurity. He has expertise in developing and managing prediction models for large-scale medical data and images. Hembroff is director of the Institute of Computing and Cybersystems’s (ICC) Center for Cybersecurity (CyberS) and a member of the ICC’s Center for Biocomputing and Digital Health (BDH).
Funding for this project is through the Michigan Sequencing Academic Partnership for Public Health Innovation and Response (MI-SAPPHIRE), a Centers for Disease Control and Prevention (CDC) Epidemiology and Laboratory Capacity grant received by the Michigan Department of Health and Human Services (MDHHS). MI-SAPPHIRE activities include sequence generation and analysis, such as sample collection and sequencing; data processing, storage and sharing; and data interpretation and analytics.
Read a Michigan Tech News article about the award.
Read a State of Michigan press release about the award.
For more information on genetic sequencing, visit the Centers for Disease Control website.