By Lynn McCain 

The paths to pathology are as varied as the members taking them. Sometimes, the path is very direct—the student knows exactly what they want to do from childhood. Other times, the path is non-linear, as is the case with Dr. Daniel Hovelson, director of informatics in the Division of Diagnostic Genetics and Genomics (DGG) in the University of Michigan's Department of Pathology.

“I actually started out studying psychology in undergrad and ended up doing some research work both abroad and then here in the United States on traumatized refugee populations. One of the things I realized pretty quickly is the only people that really knew what was going on on the research side ultimately ended up being the people closest to the data and that happened to be statisticians. I befriended them, taught myself a little bit of programming, and developed a strong appreciation for the utility of leveraging computer science and statistics to better explore various aspects of human behavior.”

This led Hovelson to the University of Michigan School of Public Health where he pursued a graduate degree in biostatistics. “I was working with a former U-M faculty named Goncalo Abecasis, whose lab was beginning to perform massive amounts of whole exome and whole genome sequencing.” In this lab, Hovelson had to build and blend his skills in computer programming, statistics and biology, which he really enjoyed. Upon completion of his master’s degree, Hovelson wanted to pursue ongoing graduate studies, but wanted to be closer to the patients. “In my pursuit of labs for my PhD, I was looking to be in a translational setting and Dr. Scott Tomlins’ lab provided that in spades. Dr. Tomlins was focused on the design and validation of innovative, clinically relevant next-generation (NGS)-based assays, and the opportunity to develop experience and expertise in this domain was quite appealing. Further, the ability to leverage these assays for interrogating unique disease cohorts and exploring novel clinical applications was exciting.”

In the Tomlins lab, Hovelson learned how to think creatively about identifying diagnostic and therapeutically relevant genetic alterations and how to translate new discoveries into potentially clinically useful assays or tools. “The domain of bioinformatics, or cancer bioinformatics, requires an appreciation for data science and statistical principles, general computer science and programming aptitude, and a deep understanding of the biology that is relevant to the problem you are trying to solve.” However, other than introductory biology courses in undergrad, Hovelson did not have a significant biology background. “I had quite a bit of molecular biology to learn to be useful in the lab – and I’m definitely still learning!”

After completing his PhD, Hovelson transitioned to industry and once again found he needed to teach himself new skills. “One of the things they don’t teach you a lot about in biomedical sciences programs is high quality software development. In order to build the types of informatics tools and workflows required to support high-throughput molecular testing at scale, a solid software development skillset and engineering team is critical. In a commercial high-throughput testing environment, I had the privilege of learning both the art and science of software development from a really talented group of software engineers. Without the sort of robust, reliable software tools we partnered to build, scaling these high-throughput clinical assays becomes extremely challenging.” Given the size of data files generated by high-end sequencer runs (typically terabytes per run) and scale of high-throughput NGS testing laboratories envisioned at DGG, data storage infrastructure must be sound, software tooling must be secure, and the ability to extract and manipulate the data to obtain usable results must be facile.

In July 2024, Hovelson was recruited to the Department of Pathology by Dr. Annette Kim, the new Director of DGG, who commented, “We were incredibly lucky to be able to recruit Dr. Dan Hovelson to our institution. He brought his keen intellect and unique understanding of informatics in a clinical production environment from his time at Strata that he is bringing to bear on our efforts to build, from the ground up, molecular informatics. Where once there was a myriad of different commercial solutions, occasionally stitched together with legacy code, Dr. Hovelson is leading a team to re-envision a modern informatics ecosystem in DGG. He is a consummate team builder, and we have all enjoyed learning from him and working with him during his short time here. I am personally amazed at the team he is building.”

As the new director of informatics in the DGG, Hovelson is focused on two major projects. The first is new software development to support the new assays the department plans to bring online, both in somatic (acquired mutations) oncology testing and in whole genome sequencing from the germline (inherited mutations) side. In this process, Hovelson is interfacing with stakeholders across the institution, other DGG leaders, and working alongside “two very talented informatics colleagues, Drs. Robert Bell and Suguna Narayan to help design and build the systems we need to support clinical execution and operation for those new assays. It is refreshing to have such thoughtful, like-minded faculty to work with every day as we design and build.” Hovelson continued, “Additionally, it has been really fun for me, especially with my bioinformatics background, to work with Dr. Marcin Cieslik. He has been an integral part of what we’re doing. Our team of informatics faculty bring many complementary skills, and I’ve learned a great deal from all involved during my short time in the division.”

The second major project is leveling up the informatics ecosystem for the legacy assays. The department has many assays that are being run and the tools supporting those assays require maintenance support. “We are thinking hard about how to optimize those systems; how to ensure that DGG stakeholders have a bit more visibility into sample tracking data processing for assays we are running today.” In addition, Hovelson is teaching, handling administrative tasks and working with informatics fellows and other faculty members.

Looking to the future, Hovelson has a couple goals. “One is for DGG to be successful. I want to ensure that we have a really solid foundation. That means software engineers, bioinformaticists, and scientists being brought onto the team who are capable of helping us build robust, maintainable software and workflows. The second is to activate the two new assays we have targeted – whole genome sequencing and DGG Oncoseq. I want to see that through to formal activation. We are still a ways away from turning those things on clinically, but I want to be sure that we have built the type of informatics ecosystem that enables us to both clinically run those assays and facilitate interdisciplinary research endeavors.” The data generated from running these assays could be beneficial for clinical research and Hovelson hopes to build collaborations with other Michigan Medicine researchers to make discoveries that could improve patient care and outcomes. Being a part of a clinical pathology department, “we have variable data types that can be compared, contrasted, or used together to enhance our understanding of the underlying biology and clinical utility. I am trying to help us design data ecosystems that can support both clinical care and discovery.” Hovelson also hopes to pursue potential research opportunities himself. “There are interesting opportunities with the digital pathology initiative here in the Department of Pathology and across the Michigan Medicine institution leveraging long-read NGS technologies. There are also opportunities to share the types of software and ecosystems we are building outside the institution. I would like to see us moving toward sharing of code bases, sharing of tools. This will make the broader biomedical and academic ecosystem richer to have those contributions shared externally.

Hovelson also wants to spend time focusing on trainees, both residents and fellows, to help them build the requisite informatics expertise that they will need for their future endeavors. “I want our trainees to feel like they are leaving having built up enough familiarity or even depth of expertise around informatics, topics that will be relevant for where they are headed. As trainees pass through the DGG, I want to give them additional tools that can help them be better pathologists and team science collaborators as they move forward. I look forward to working with Suguna, Robert, Ul (Balis), and others to ensure our informatics interested trainees can get the training they need moving forward. Technology moves very fast with the emergence of AI into all aspects of daily life - including molecular pathology. It would be a disservice to trainees not to at least give them some high-level experience and some hands-on experience with these technologies.”

Outside of work, Hovelson delights in spending time with his family. He has four children between the ages of 12 and 3, with his wife Sonia. “They keep us very busy. I enjoy spending time with them and planning how to travel economically with a big crew. The combination of my growing family and finding new places to adventure brings me a great deal of joy.” Hovelson has one son, followed by three daughters. “It is a combination of sports and Taylor Swift in our house.” His wife, Dr. Sonia Hovelson, who trained at Michigan Medicine as an Ob/Gyn, is now working for a utilization management company reviewing insurance claims. “She is easily twice as intelligent as I am. Often we discuss drug approvals or treatment approvals that are squarely in the oncology or pathology space. She keeps me on my toes.”