David Gibbs

What has been your favorite project that you worked on?

Gibbs: I think my favorite project probably would’ve been one that finished up a few years ago. It was at the very end of 10 years of projects that were called the TCGA [The Cancer Genome Atlas] that covered 32 different kinds of cancer over the years. At the very end, they decided they wanted to do a PAM cancer project. This type of project looks at the similarities and major differences across every kind of cancer, like what can be learned from looking at all tissues at the same time and then also are there things that stand out between certain tissue groups that are unique. The interesting thing was it was unfunded. It was an all volunteer project that took at least two years. There were so many people and we had weekly meetings. I worked on the cancer immune landscape paper which looked at how cancer and the immune system interact. That was where I learned a lot about the immune system and how it operates while a tumor is growing. It was a place where we learned a lot and gained collaborators that we still work with. At the end, it was published in Cell and was this huge package where all of the papers were published at the same time and it got its own webpage. It was a really big deal. We all flew to Washington DC when it was published, gave presentations and met everyone involved in the other projects. It was really cool; it felt like a community event. We still work with those people even years later and it has even turned into one of the projects in our lab called the iAtlas. That’s directly where it came from. 

STEM and branches of computer sciences have so many subdivisions and specializations nowadays. For someone who is generally interested in the computer sciences, do you have advice for how to go in?

Gibbs: Well, you are talking to the wrong guy because I got into a lot of trouble with that myself. I really liked everything and looked at everything. I tried to do everything and feel like I haven’t gone into a very narrow slice. But, I think in bioinformatics you can get away with that because you are often collaborating with people who have really deep domain expertise. So maybe that’s my expertise – knowing how to do an analysis in general. A lot of times the analysis is a collection of general methods that can be applied anywhere, so you just need someone with the biological knowledge to point you in the right direction. That’s the way I worked a lot. 

Here is another thought: It is important to be able to pick something you are excited by or at least like, because if you start a project on it, it takes a really long time to complete. So, you want to pick something that doesn’t drive you crazy. I think that’s probably most important – picking a topic that you enjoy studying.

Why did you pursue a PhD?

Gibbs: I didn’t initially intend to get a PhD. I was a post-Bacc student. I worked in chemistry labs and was a post-Bacc student in computer science because I wanted to do something in that direction. While I was doing that, I realized there was this program that sent you from the computer science department to the masters program, so I decided to do that. I was a little bit of an older student and had been working a while, so I was really appreciative of being in school again. I talked to the professors often, so they got to know me quite well. It was when Obama was president and it was right after the 2008 financial crash, and he was doing these big government funding initiatives. During this time the school received quite a bit of funding for students and so, the school approached me about getting a PhD.  I wasn’t in the place when they asked me but they said if you are, apply right now. Like, this week. So I said, “okay” and I did. I applied right away and got in immediately. It was really sudden. Later on, when you are halfway through the PhD program you have to do the Quals, or the qualifying exams to be able to proceed into the next half of the PhD. At the Quals, one of the professors asked me, “so why did you want to get a PhD?” I felt like I didn’t have a good answer. It was something along the lines of “I enjoy the learning process so much and really enjoy the research side. I wanted to continue doing research”. 

How many years of school did it add?

Gibbs: My PhD portion took 4 years but before that I did 3 years of post-Bacc computer science plus 4 years of chemistry. I also completed several years of postDoc, so about 11 years of school. In the PhD program, you get paid. You get a stipend and your tuition is paid for. A lot of times they have you work for it like teaching as a TA but usually you receive a stipend. 

We were wondering, since you are a senior researcher at ISB, what is your biggest piece of advice for someone going into research?

Gibbs: It sounds really like a hallmark card, but my advice is to have enough willpower and not give up when you hit blocks. You will see the smartest kids in the world and they have been able to roll through anything without studying. Eventually everyone hits a wall and the people who don’t give up right away are the ones that go on.  It doesn’t matter how smart you are, you have to be reasonably smart, but it is much more willpower and stamina to get through something. That’s the deciding factor. Some kids will give up or there are people who are afraid of failing so they don’t try something new and that holds them back. You have to stick with the problem and try things out. Eventually you will make progress and get past that. That’s pretty much it. Figuring a way around the problem and not giving up when it seems new or hard. Also, whenever I’m stuck, going to the literature always helps. I don’t do it nearly enough. But, whenever you’re stuck, you can go to the literature and get some ideas. Seeing what other people did and how they approached a problem, even if it is something unrelated to your area. You wonder if it could work for your problem. A lot of times that is what breakthroughs are: people borrowing ideas from other fields and making progress.  It happens a lot. 

If you could pursue a career in any other field what would it be?

Gibbs: Statistics. I almost wish I would’ve gone back into statistics. I really enjoy it and it’s one of those things where you have a deep knowledge of something and get to apply it to all these different areas and learn about what they are doing. I still fantasize about going back to a masters program. I have even looked it up. UW has a masters program about statistics. I would totally do it if I had time. 

Is informatics part of statistics?

Gibbs: Yeah, it uses a lot of statistics. It isn’t just statistics. It is really statistical modeling. 

I feel like beginner statistics is number crunching into the calculator. It seems really cool when you get to the higher levels of it, though. 

Gibbs: Yeah, it really is problem solving and you can’t just…Well you can reuse stuff but each problem must be tailored a little bit. You have to kind of tailor your solution to what it is. It’s cool and now, it is super computational.  I really liked this program where you write a computer program that gets compiled and then that is your statistical kernel that does all the computation. It is writing computer programs now. 

What is the dumbest way you have been injured?

Gibbs: I am a little bit clumsy. I broke this wrist and got a plate in it skateboarding. I used to skateboard a lot so I wasn’t terrible and was at the skatepark towards the end and was just turning on the ramp. I don’t know what happened, but I just broke my wrist on nothing. It was definitely bent. They call it a spoon [because of how it bends]. 

[Briefly returning to the subject of graduate school,] When I was looking at graduate programs, I just chose whatever was convenient. I think what people should do instead is look closely at the labs and schools you are considering.  It ends up making a real difference down the road. I got kind of lucky and ended up in a good lab, but it was a lot of luck on my part. It is a way better idea to know what you are getting into and research the lab. There’s a lot of ways you can figure out whether a lab seems like a good fit. Is this a boss I would like to have for a few years? Talk to the people in the lab and get a sense of how they are doing.