James Graham Brown Cancer Center

DataseamGrid helps UofL researchers identify 30 potential drugs to fight COVID-19

LOUISVILLE, Ky. – The novel coronavirus may have K-12 students in Kentucky’s school districts learning at home, but researchers at the University of Louisville are using the computing power of thousands of computers in classrooms across the state to identify drugs to treat COVID-19. The desktop computers are part of the DataseamGrid, a network of computers housed in classrooms of 48 Kentucky school districts as part of a partnership designed to support research, education and workforce development.

John Trent, Ph.D., deputy director of basic and translational research at the UofL Health – James Graham Brown Cancer Center, conducts virtual screening to discover new cancer drugs using the DataseamGrid for high-volume computations. Today, he has the computers at work 24/7 to identify the most promising drugs and compounds to fight SARS-CoV-2 and its disease, COVID-19.

“In these unprecedented times, we had a resource where we could potentially make an impact quickly and switch over from some of our cancer targets to SARS-CoV-2 targets,” Trent said. “We have been very successful in doing this in cancer for 15 years. We are using the same approach in targeting the coronavirus, just targeting a different protein.”

Established in 2003, Dataseam is funded by the Kentucky General Assembly to provide computing infrastructure, workforce development and educational opportunities for students and staff in Kentucky school districts. Available computing power in those units is put to work performing computer modeling calculations to screen anti-cancer drugs for Trent’s team and collaborators at UofL.

“Like a lot of industries, we have shifted our skills and infrastructure to address this issue,” said Brian Gupton, CEO of Dataseam. “We are always going to have cancer, but at least for the time being, we are glad the DataseamGrid is here for Dr. Trent to screen those drugs.”

In mid-March, Trent and his team entered new data onto the DataseamGrid, along with UofL’s dedicated research computers, in a two-pronged approach to match three-dimensional models of proteins in SARS-CoV-2 to drugs and compounds that could help in treating or preventing COVID-19. The DataseamGrid provides up to 80 percent of the computational power for these projects.

The first approach is to test about 2,000 drugs already on the market and another 9,000 investigational drugs and nutraceuticals that have been tested for toxicity to isolate those most likely to be effective against the virus.

“For the immediate approach, we are testing drugs that already are approved by the FDA or have been tested in humans. If we find activity with those drugs, we could get them into patient trials a lot quicker,” Trent said. “However, these drugs obviously were designed for something else and they may not have the same efficacy of a very selective drug.”

To find that selective drug, Trent’s second prong of research includes computational models to screen 37 million small molecules and compounds against the target proteins in SARS-CoV-2. These molecules could be used to develop a new drug specifically to treat the virus. That process would take more time, however, to obtain FDA approval.

“That initial discovery of a new, more-selective agent is more long term. You are looking at 12 to 18 months before you would even think about testing those in a patient,” Trent said. “But time is of essence at the moment, so we are doing both things at the same time.”

Using the DataseamGrid and UofL research computers, Trent and his team are screening the drugs and small molecules against 3-D structures of four proteins in the virus to see which compounds might bind with the proteins. A drug that interferes with the activity of any of these proteins would reduce the virus’s ability to spread.

Trent began the research with the first two proteins described for SARS-CoV-2: the main protease, an essential enzyme used by the virus to break down viral proteins and make new virus particles, and spike proteins, the triangular “knobs” the virus uses to attach itself to host cells. These spikes are the knobs commonly seen in graphic images on the surface of the virus. Trent now also is testing drugs against two additional target proteins that were described very recently.

So far, the process has identified about 30 drugs as potentially effective against SARS-CoV-2. Trent recommended these for biological testing by other UofL researchers in the UofL Center for Predictive Medicine for Biodefense and Emerging Infectious Diseases (CPM). Directed by Kenneth Palmer, Ph.D., the CPM is one of only a few labs in the United States capable of testing the drugs against the virus. That testing is expected to begin in mid-April.

If the CPM researchers find the drugs to be effective against SARS-CoV-2 in the lab, they could be moved to the next phase of testing in animal models, testing that also may be conducted at CPM.

“This computer modeling is an excellent way to identify the best potential candidates for laboratory testing rapidly, and this strategy could lead to relief sooner rather than later for patients suffering from COVID-19,” Palmer said.

Gupton says it is good to know the DataseamGrid continues to support urgent medical research even though students are working from home.

“Ironically, we hope to return to cancer research as soon as possible,” Gupton said. “Even though the students are not in classrooms, Kentucky school districts are providing them with instruction, technology, internet access and even meals. The districts’ Dataseam systems engineers are supporting both the university’s work and K-12 efforts. We all are proud to be ‘digital first responders’ in Kentucky’s part of the global fight.”

As part of the Dataseam partnership, UofL provides academic scholarships annually for students from participating school districts who come to the university to pursue a degree in science, technology, engineering or math.

By Betty Coffman, betty.coffman@louisville.edu

Build Your Dream at the University of Louisville

(Louisville, KY) – Thirty-two high school students and parents from Russell Independent, Paducah Tilghman, Elliott County and Whitley County Schools toured the main campus, cancer center and engineering maker space at the University of Louisville to learn about areas of study, career options and scholarship opportunities.

The full day featured visits to the research labs at the James Graham Brown Cancer Center where Dr. John Trent and several teams of scientists use the data processed on the DataseamGrid to search for and develop potential cancer drugs.  The multi-discipline approach employs physicians, engineers, chemists, biologists, computer science specialists and others to fight cancer.

Representatives from the J.B. Speed School of Engineering outlined degrees available and highlighted work experience each student receives through internships at local companies.  Students graduate in five years with a Master of Engineering degree and marketable work experience that takes them beyond the books.

Students were really impressed with the associated 1B First Build professional maker space where students can co-create potential products.   Backed by GE Appliances, the campus facility provides students access to world-class engineering and design talent as well as some of the latest manufacturing equipment.

Students from schools in the Dataseam program interested in studying Engineering, Science, Medical Research or Health Sciences should apply today (link).  Specific DataseamScholarships as well as other programs to help with the cost of college are available.  Deadline January 15.

Partnering universities have provided over $2.2 million in 4-year scholarships to students from Dataseam-participating schools to advance Kentucky’s workforce in STEM and STEM education.

One Thousand Years of Cancer Research

In the first 70 days of the calendar year, the DataseamGrid completed over 1,000 years of critical cancer research.  Since January 1, 2017 researchers at the James Graham Brown Cancer Center have used this massive Kentucky cloud-based computing system to investigate tumor suppressor pathways, target specific genes and identify existing drugs that may be used to fight different cancers.

This unique computing system provides 40,000 Gigahertz of processing power from desktop computers sitting in 38 school systems across the state.  It extracts the maximum amount of computing power from every system around the clock, not only during evenings and weekends, but also between classes and other low use periods.

As one of the largest distributed computing systems in the world, it supplies twice the raw processing power of dedicated super computing at the University of Kentucky and University of Louisville combined[1].  If you purchased that much power from traditional cloud-based services, the state would have paid nearly $500,000 since early January[2].

Teams of Kentucky-based researchers have been looking at over 130 targets against their database of 25 million potential compounds to identify potential drugs.  They have also extended their search efforts to scan national databases of thousands of drugs including FDA approved drugs, to determine if some of those drugs might be useful against other types of cancers and different treatment programs.

The potential to repurpose existing medicines or substances may help get the benefits to people more quickly.  It is an area of focus for the NIH (National Institutes of Health) and University of Louisville cancer research teams are concentrating efforts in this area.

Supported and maintained across 38 school districts from Pikeville and Martin County in the east to Union and Crittenden Counties in the west, these 12,000 plus school computers take on these research requests that might consist of 50,000 individual tasks, with each task taking many hours to complete.  If local equipment, network or power conditions put a school or school system offline for any time, the system dynamically redirects tasks to other available computers so the important research marches on 24-hours a day, seven days a week.

Dataseam developed and manages the DataseamGrid in collaboration with 38 Kentucky school districts. Dataseam also works with local school districts, universities and academies to promote STEM educated workforce, particularly in bio-medical research and aerospace industries.  Over 7,000 teachers and 90 technicians participate in training and industry-based certification programs as part of this effort.

[1] Computing power based on number of intel cores as documented at https://www.rc.uky.edu/high-performance-computing/hpc-hardware/ and http://louisville.edu/it/departments/research/cyberinfrastructure

[2] Calculated using http://calculator.s3.amazonaws.com