Second annual ATOM summer training program empowers and equips students

Six graduate students from Butler University and Nova Southeastern University completed a virtual 10-week summer training program with the Accelerating Opportunities for Therapeutics in Medicine (ATOM) Consortium at the end of July. Their efforts are the latest part of ATOM’s mission to put accelerated, computer-based pharmaceutical discovery in the hands of the scientific community.   

The ATOM Summer Training Program is an essential part of this democratization goal. It equips future drug developers and pharmacologists with expertise in data science, AI, and drug discovery. 

“ATOM is not only about the technology but putting tomorrow’s technology into the hands of the workforce today,” said Eric Stahlberg, Ph.D., ATOM co-lead and director of Biomedical Informatics and Data Science at the Frederick National Laboratory. 

ATOM is a collaboration between national laboratories, universities, and the pharmaceutical industry to combine scientific studies, data collection, and computing to shorten the research and development time for new medicines. By combining disciplines, the consortium is working to speed up the traditionally long and expensive process. 

An integrated opportunity 

During the fast-paced program, the students learned to code and to create computer-based models to predict whether certain chemical compounds would lead to cell death or pass into the brain to possibly attack brain cancer cells. The method embraced machine learning models and models based on fundamental mechanisms of biology. (Machine learning is a method that uses data to train algorithms to make predictions.) 

The students, five Pharm.D. candidates and one Ph.D. candidate, presented their final projects via web conference to scientists from Frederick National Laboratory, Lawrence Livermore National Laboratory, ATOM headquarters in San Francisco, Butler University, and Nova Southeastern University. 

The program aimed to teach the students how to write code, perform machine learning, analyze models, and understand assay development. They built the models using the open-source RuleBender software and AMPL (ATOM Modeling PipeLine) packages. Their models and the abundance of analyzed data advance ATOM’s modeling and molecular design program.  

The students worked on their projects under the lead mentorship of Amanda Paulson, Ph.D., data science ATOM fellow at the Frederick National Laboratory, along with Susan Mertins, Ph.D., visiting scientist at Frederick National Laboratory and assistant professor of science at Mount St. Mary’s University. 

“This year, we developed a set of interrelated projects. Each model is useful on its own. However, the models may be used in tandem to design new possible drugs that achieve multiple goals at once,” Paulson said. “For example, we will try to design drugs that not only pass into the brain, but also lead to cell death for cancer cells once they are in the right location.”  

Building the future workforce 

In this year’s expanded program, Paulson created a virtual experience that included seminars from ATOM researchers and career talks from Pharm.D. professionals in the data science field. The approach let the trainees to connect with each other and network with researchers from California to Florida. 

Screenshot of a virtual meeting
Trainees, mentors, and supporting scientists and staff gathered virtually for the final presentations. 

Caleb Class, Ph.D., assistant professor of pharmaceutical sciences at Butler University, led a two-week Python Programming Language boot camp for all the summer trainees in May. In the remaining eight weeks, students built, tested, and refined their models. 

Zahra Mehrabi, a Pharm.D. candidate at Nova Southeastern University, equated it to “creating something out of nothing.” Most of the students had little experience with coding as they began the project, and none had done machine learning before. 

There were technical challenges, too. Ryan Friedrich, who is pursuing a Pharm.D. and M.B.A. at Butler University, had to discard a large, otherwise useful data set because the available format made it impossible to extract crucial pieces of data. Kendra Schorr, a fellow Pharm.D. student at Butler, had one machine learning model that trained poorly because of inadequate and insufficient data. She had to remove it from her final analysis. 

“There was definitely a learning curve. … Then you have days of the week where things just click,” Schorr said during the discussion after the students’ final presentations. 

The goals were daunting, especially at first. However, working closely with their scientific mentors and the ATOM team, the students successfully finished their models. They were thrilled with what they were able to achieve in such a short time. 

“I wish we had more time. I had fun,” Mehrabi said. “Now, I wish we could do more.” 

Jim Brase, M.S., ATOM co-lead and deputy director for Computation at Lawrence Livermore National Laboratory, told the students that training the community will remain one of ATOM’s missions. Frederick National Laboratory’s Stahlberg agreed. 

“ATOM is building the future workforce and integrating the disciplines,” he said. 

‘From nothing to having so much’ 

The students spoke to the value of the opportunity after the presentations. Sarah Abu-Salih, a Pharm.D. candidate at Butler University, appreciated “going from nothing to having so much” and “seeing experiments actually work.” Friedrich learned to troubleshoot complex analytical and coding problems with just the resources at his fingertips. Caiden Lukan, a Pharm.D. candidate at Butler, said he built on his prior coding experience and learned new aspects of data science he plans to use in his career. 

“We are very proud of the trainees’ achievements this summer, and excited to build upon the foundations of their work to reach new frontiers in computational drug discovery,” mentor Paulson said. 

Natalia Noto, who is pursuing a Ph.D. at Nova Southeastern University, summed up the experience. 

“I am capable of doing a lot in a very short time. I learned I know more than I thought I did,” she said. 

Speed, learning, and capability in the sciences are at the core ATOM’s primary aims, for both humans and machines. If Noto’s words are any indicator, both benefitted from the program.