Building a (bioinformatics) Dream Team
Over the course of a long career in bioinformatics at a National Laboratory I have had the fortune to build and evolve 2 great teams. One team was part of Human Genome Project from its pre-inception in 1987 until the publishing of the first human reference genomes in 2000. The other was a pathogen bioinformatics team that I began building in 1999 and just turned over to my successors when I retired recently. With the benefit of 20/20 hindsight it seems interesting to look back on how these teams came about and what I learned along the way.
The first admission is that sheer luck plays an enormous role in team building: the intersection of my having funding for a position and a suitable candidate being available is often a miraculous window of opportunity that normally is stuck closed. In 2004 I desperately needed a good bio-statistician after my former one had retired. Out of the blue, I heard from a student who I had taught in 1998 when I moonlighted for UC Berkeley Extension and taught the first bioinformatics class in the San Francisco Bay Area. I had 55 students, some who drove down from UC Davis or up from UC Santa Cruz or Stanford. The very best of the students was working for Oracle at that time, but in 2004 he was part of the West Coast branch of a biotech company that was suddenly shut down. He got his PhD in biostatistics while working for me full time over 7 years, and his brilliant work continues today. Another example was when a member of my Human Genome team asked if I had any part-time work for a biologist who was finishing her PhD at UC San Francisco (and was renting a garage to live in and babysitting to supplement her meager student income.) I found her some part-time work but after she completed her PhD I hired her to prop up my massive shortcomings of biological knowledge; she helped me succeed on multiple pathogen genomics projects and is still contributing to the team today while working part-time remotely.
Of course, sometimes you have to make your own luck; in 2007 after nearly a year of failing to find a suitable bioinformatics postdoc I sent emails to all the academics I knew. Steven Salzberg at U. Maryland wrote back that he had a person finishing his PhD who grew up in Berkeley and who might love to move back to the Bay area. That’s how I found my successor for the CS/bioinformatics side of my team.
You can’t always start a team from scratch, of course. Sometimes you inherit an existing team, or, as happened to me on the Human Genome Project, you get people dumped on you who you might never have selected on your own. Biology was treated as a third-rate discipline at my former National Laboratory until the early 2000s; through much of the 1990s my bioinformatics group was the designated dumping ground for people who weren’t good fits on the various first-class programs (lasers, nuclear weapons, etc.) I would typically get a call about 4:30pm on a Friday, telling me that so-and-so would be showing up to work on my team on Monday morning; nobody even asked me if I had funding, space, or if the skill set of the individual matched what my team did.
This was quite a managerial challenge. Being in computer science, this meant that I got people on the autism and bipolar spectrums dropped on me. Being in the SF Bay area, it also meant that I got to work with people whose gender alignment or reassignment was too much for their former colleagues to handle with grace or even professionalism. There were also a few who were incompetent or unwilling to work hard and were just passed on for me to either find a way to salvage their careers or let them go.
Through all of this, I learned that I absolutely had to treat each employee as a completely unique individual and learn what their strengths, weaknesses, motivations, passions, and fears were. I had to figure out how to pair each newcomer to an existing team member to give them the best chance to learn quickly and fit in. This usually meant I had to determine initial tasking and peer support so that they did not fail on their first assignment. It also meant that I had to be personally committed to their success and always be aware that square pegs never fit in round holes no matter how hard you pound. Amazingly, the success rate was phenomenal; on my last day before retiring, one of those folks who was beamed in on a Monday morning came to thank me for saving his career by giving him a literal “last chance”. Twenty years later, his career in genomic analysis is still going strong.
I also learned that I had to sometimes take risks on people who would have been easy to have passed up. In 2000 a population biologist on a Lawrence fellowship was mooching a lot of my Unix computer time, running some modeling code written in a “dead language” (FORTRAN). Although her constant usage of my computers was somewhat annoying, she would always graciously throttle down whenever it was interfering with my team’s production work. In the summer of 2001 she walked into my office and said “Hey, Tom, you guys are doing some cool stuff with pathogen DNA analysis. I have one year left on my fellowship and I am bored with what I’ve been doing for the past 2 years; can I work for no cost to you for one year?”
I must confess that my first thoughts were along the lines of: “She’s a population biologist who programs in FORTRAN with no DNA analysis experience at all; maybe she will just slow down people on my team while she learns?” But my gut instinct was to take on anyone who was bright enough to win a coveted Lawrence fellowship and dumb enough to want to work for me, despite the apparent mismatch of training and experience. Barely a month later, the 9/11 attacks occurred, and my team was swamped with pathogen signature work as we responded to the subsequent anthrax letter attacks and helping the CDC to create better PCR signatures. She quickly became the absolute superstar of my team and remained that way until cancer killed her in 2016.
Luck may bring good people to join your team, but it is the establishment and maintenance of a good culture that lets your team thrive and survive tough challenges. We will investigate that aspect in a future post.
(Tom Slezak was a student employee at Lawrence Livermore National Lab from 1975-77 and a full-time employee from late 1978 until September 2018. He is now the CEO and Co-founder of KPATH Scientific, LLC, providing technical and managerial consulting to several biotech startups.)