PhD Reflections III (The End)

A hodgepodge of things I haven’t mentioned yet and thinking about what’s next

While this post is ostensibly about the final portion of grad school, much of it covers critically influential aspects of grad school that didn’t fit cleanly into the first two posts.

I jump from “How to Learn to Code” to “The Trailer” to “Departmental Retreats”, all against the backdrop of trying to decide what I want to do after I graduate.

4th year, Spring Semester (2018)

Half of my brain was again on the startup class this semester. But the semester was also incredibly productive in terms of research. There were two reasons for this productivity. First, Mauro and I had our TODO list. TODO lists are a rare and wonderful commodity in research. We had gotten back review comments from Nature Genetics. The comments were extensive yet favorable. We went through each individual remark and noted every addressable action item we could find. That became our TODO list. I handled the items that required more analysis; Mauro tackled the ones that were restricted to writing. Having a checklist arguably indicates you aren’t even doing research; either way, it made things much easier.

The second reason I was so efficient was because I was under a strict deadline…

“Well, crap. I actually gotta tell him.” That was my first thought when I read the email from Google telling me I needed my PI’s contact information. When your paper is in the middle of the review process and it’s the lab’s first paper, telling your PI that you want to disappear for four months is intimidating. It was one of the only hard conversations I had to have in those five years. When I finally steeled myself, went to tell Mauro, and finished explaining the situation, his response was:

As a person, I’m really excited for you. As a PI, this is really hard.

Flawless. And catchy.

PIs and Grad Students

If you’re reading these posts to learn about graduate school, this is an important section.

Mentor/student relationships in grad school are notoriously difficult. I’ve read a lot of articles complaining about this issue, but few seem to understand exactly why the relationship is complicated. There are three reasons, which I’ll list in order from the most general in working relationships to the most specific to mentor/student:

1. The power ratio is highly skewed. This is the case in most boss/employee relationships. Here, however, it’s worse than average. Beyond the valid complaints that PIs have excessive control over when students can defend, they also control recommendations. In most companies, if you’d rather not get a recommendation from your direct boss, you likely have other people you can use as references (e.g., a project manager instead of your technical lead). Graduate students about to enter the workforce are always going to be asked for their PI’s contact information.
2. To stereotype, PIs are intelligent, they’re driven, and they love science. PIs are PIs because they were excellent students, then excellent graduate students, then excellent postdocs. They were the best in their field at doing research, and they did research for a long time. Then, they got a professorship. As professors, they are immediately expected to stop doing research themselves and start managing graduate students and postdocs who are doing research in their stead. Doing good research yourself is literally worse than useless for being a good manager of other scientists. At best, they’re orthogonal skills. To continue stereotyping, it turns out that the jokes about scientists and engineers being poor at people skills are largely based in reality. We’ve ended up with a system where the selection criteria for who should become a PI has little to do with the day-to-day realities of how PIs are expected to spend their time. To summarize, most PIs are naturally bad managers who have minimal experience managing. A trivially obvious consequence of this is that the graduate students the PIs are managing don’t enjoy the experience.
   • As an aside, this is in no way an accusation against PIs. It’s the system that’s broken. The silver lining is that it seems like more effort is going in to providing PIs with managerial training.
3. More than anything else, the relationship is strained by misaligned incentives. PIs are rewarded for a certain set of circumstances, grad students for another. There are several ways this plays out, but an easy example to explain is how older graduate students are more valuable to a PI. A PhD is a training environment; when students start grad school, they know close to nothing. The student’s PI (and other members of the lab) spend an enormous amount of resources, primarily in the form of time, educating new students on how to conduct high-quality research. A few years later, when the student has finally started to become independent, they’re ready to leave. A PI can’t help but think, “Wait a minute. If I can get this student to stick around for another year or two, maybe I can get another paper out of them. After all, I’ve put all this time and energy into them.” It’s a logical thought, and not particularly sinister. Industry does this, too. The difference is that industry incentivizes an employee to stay by offering them a raise each year or a promotion after a few years. PIs have neither of these options. Because they have no reward mechanisms at their disposal, they only have threats. The threats aren’t explicit, they’re subtle. Academia is full of circulating stories of PIs who imply that students need to finish “one more paper” if they want to defend successfully or find a prestigious job. In reality, that extra paper is usually significantly less important to the outcome of a job application than a ringing endorsement from the PI. Personally, I never ran into this problem. I did struggle some with the fact that PIs are rewarded for students staying in academia, while most students (myself included) are necessarily going to end up elsewhere. Even if you assume the number of professorships double each generation, each PI usually trains dozens of students in their lifetime. PIs want students to do nothing but focus on writing papers; students want a holistic education. For me, learning a second programming language like R might have roughly doubled the number of job prospects I would have in the future, but it would have zero net effect on the measured productivity of Mauro’s lab. On the other hand, if I had produced a second paper, that would have doubled the measured productivity for Mauro and likely not mattered one iota for any of the industry jobs I wound up applying for. How should we have come to an agreement on how to spend my time when the incentive structures don’t match?
   • As a second aside, this bullet point is slightly an accusation against PIs. But it’s primarily still an accusation against the system. Humans aren’t built for altruistically fighting against institutionalized incentives. Some PIs are much better than others, but the root of the problem is still the underlying incentive structure.

How to Learn to Code

How to Learn to Code marked my favorite weeks of each summer. The class spun out of a couple of students struggling with the difficulty of simultaneously learning structural biology and Python during Structural Bioinformatics our first year. It grew into a campus-wide organization, and I was lucky enough to get to teach all four years.

TODO LINK

Even though I was starting an internship with Google in summer of 2018, I still got a couple of weeks of teaching in before heading to Canada. The rest of that year’s class was taught by Kimiko. She and I managed to find a reasonable amount of time to prep the full curriculum for the class before I left. Having explicit time set aside for teaching is one of the things I miss most about grad school.

5th year, Fall Semester (2018)

I spent approximately four months away from grad school interning in Waterloo, Ontario. It was the definition of a career-changing experience. Actually, working there was surprisingly similar to grad school. My team was full of smart, talented people who were interested in building useful tools for the life science community, just like my colleagues in grad school.

To be honest, the “career-changing experience” part was the line item on my resume. While I was interviewing, I literally had a person tell me that a large reason I got the interview was because Google was on my resume. It was an awkward but insightful moment for me.

I also don’t want to downplay my time there. The summer was full of new experiences, both technically and generally. I would advise anyone doing a PhD to spend a month or four in an industry internship, even if they’re planning on doing a postdoc. The perspective is worth the delay.

Done

Coming back from Waterloo wasn’t great. I had a continual sense of regression—like replaying an already beaten level in a video game—which I never managed to shake. Most of my friends had felt like they were ready to graduate for years, so I didn’t get a lot of sympathy.

Mauro did his best to motivate me. I was working with Dan, a new graduate student in our lab, to write both a book chapter we’d been invited to publish and potentially a second paper. There were days when I truly believed we’d be able to get it done in a year. But those days never carried into weeks.

Within a few short months, I started panicking about Google jerking me around about a job offer. Every few weeks they would call me and tell me that they needed a few more weeks to decide about my situation. By the beginning of December, I decided I needed to get other offers. I spent a lot of time talking about it at the trailer.

The Trailer

Most days, Megan, David, and I, along with the rest of the crew (by this point, Mauro also had Dan, Rachel, and Keenan as grad students), would eat lunch at The Trailer. Technically, the establishment was called “The Courtyard Cafe”, but we never referred to it as such, given that it was literally housed in a trailer. Over the years, many good things happened at The Trailer. It was where:

• MazeDay was born
• my friend Mac and I would joke about our smart-toilet startup
• I developed my reputation for hoarding plastic cutlery
• burrito bowl dice was played (diarrhea, or not?)
• we witnessed David’s steadfast dedication to his single lunch
• Megan occasionally vented
• I became known as “the cheese guy”

Before I left for Waterloo, I started making friends with one of the servers. Maybe we were friends, but that seems a bit presumptuous given that I don’t know his name. He didn’t speak much English, so most of our interactions were based on jokes about how much I like cheese. I’d always ask for extra with my burrito bowl. He turned it into a game where he would see if he could give me too much cheese. He could. He’d usually do it while maintaining aggressively strong eye contact, and everyone in line was laughing.

When I came back from Waterloo, he was so excited that he filled a to-go container full of cheese and gave it to me as a gift.

Go to lunch with your colleagues; interact with the people around you. It’s worth your time. Make it worth their time, too.

5th year, Spring Semester (2019)

Every semester of graduate school was easier than the semester before it. That was true up until the last semester. The last semester was stressful.

It was the standard stress of life transitions. I had to figure out what I wanted to do next, I had to interview, I had to write a thesis, defend it, and generally wrap things up. The semester involved a lot of deadlines, though nothing particularly extraordinary.

Job hunting

Here are my “stats”:

• I sent out applications to 23 companies.
• I had approximately 14 phone calls, though some companies doubled up.
• I had 5 on-sites (1 in Boston, 3 in California, 1 in Durham).
• I ended up with 3 offers.

On the one hand, I want to say I’m pretty comfortable with interviews at this point and almost enjoy them. On the other hand, I smelled like stress sweat (which is an order of magnitude worse than smelling like normal sweat) after even the shortest phone interview. Job hunting is a physically and mentally draining process. Even without nerves, interviews mean 100% focus and engagement for an extended period of time.

Despite the amount of effort it took (and I skipped over the hundreds of websites I looked at to come up with those 23 applications), the “spray and pray” approach to job hunting was totally worth it. Grad school had provided me minimal exposure to the bioinformatics industry. Job searching was a phenomenal way to get a lay of the land by doing everything from reading company websites to being flown out to talk to people and tour labs. Getting to see all of the amazing work life scientists are doing all over the country was informative and inspiring. If you’re about to graduate, even if you’re pretty sure where you’re going to end up, I’d advocate for shopping around. You might be surprised.

Of course, I also thought a lot and learned a lot about job interviewing during this time. Interviewing certainly falls into the “practice makes perfect” category of activities. There are already lots of good resources about how to interview, so I’m not going to detail my thoughts on interviewing here. I will say that if you are going to interview it’s certainly worth taking the time to plan ahead. Be intentional and strategic about how you present and sell yourself.

There’s no excuse for winging an interview.

Choosing a job

With three job offers (and Google still hanging around in the background), making a final decision on an offer was its own source of stress. Many factors went into the decision; it’s overkill to get into all of them. There was one moment in particular, however, that was memorably influential.

The Durham company interview was wonderful. The company was doing amazing work; I enjoyed every individual I spoke with and could see myself making valuable contributions to the small company. The hiring lead was doing a stellar job pitching the company. He knew I already had other offers and was fairly close to making a decision, but I wanted to hear him out. Truth be told, he nearly convinced me. Accounting for cost of living, his offer was the best (by some margin). Near the end of the conversation, though, he unwittingly turned the tables on himself. He knew I was thinking about San Francisco and said:

Here’s the thing. San Francisco is always going to be there. If, for whatever reason, you decided this job isn’t for you in a few years, you can just go to San Fran. It’ll still be there. This company, though, is growing. In a few years, it’ll be impossible to get a position like this because we’ll have already filled it.

Initially, I agreed with him. After all, his logic is sound. Then, the follow-up thought derailed everything. Even though San Francisco won’t change in a few years, my ability to get there likely will. North Carolina is a good place to settle down; things like buying a home and having kids makes it exponentially harder to pack up your life and move across the country. In essence, making it to California at this point was a once-in-a-lifetime opportunity—not because California was going to move on, but because my life was. I had a mini-existential crisis in the middle of the interview.

I think about the passage of time, and the speed at which that passage occurs, often. “Life is short”, blah, blah, you’ve heard it all before. In a conversation Mauro and I had, he framed the sentiment in way that especially resonated with me:

I’ve had a fairly successful and productive scientific career. One of the reasons I’ve been successful is because we all have an internal clock, and mine ticks just a little bit faster than everyone else’s.

Since I’ve heard that line, thinking about mortality has become more visceral. “Ugh, I just wasted two hours on that goofy movie.” Tick tock, tick tock. “Do I want to commit to writing a blog?” Tick tock, tick tock. “Should I take this job opportunity in California now or take the safe, comfortable bet?” Tick tock, tick tock.

Tick tock, tick tock.

Thesis writing

Before I left for California, or any other place, I needed to do a couple other “small” things first. I needed to write a thesis. Thesis writing is one of those things that varies widely among PhD programs. Some of my friends spent months on theirs. My philosophy on the task was shaped by a quote I heard from one of my interviewers:

Whenever you’re writing, it’s important to keep your audience in mind. In the case of a thesis…that’s no one.

Mauro completely agreed with the sentiment since it left more time for research. Oddly, my mental state kicked back into high gear, and some of the coolest pieces of data I found during grad school were found in the last few weeks before submitting my dissertation. My retrospective hypothesis is that I knew I wouldn’t have to do the legwork of getting the results out the door, so I simply had fun with it instead, which allowed me to focus more than I had in the previous semester.

For two painful weeks, I pulled together all of my chapters (putting legitimate effort into the intro, conclusion, and the new research chapter). I mucked through formatting, citations, and the like. Then I submitted. I was done.

Departmental Retreats

Well, I was done with dissertation writing, which meant I had a single weekend to relax and enjoy our departmental retreat. Retreats were always one of my favorite times of the year. For me they consisted of:

• Spending all day listening to fascinating science.
• Spending all night hanging out with friends.
• Trying to win things. I’m a sucker for some competition. My first year I won $50 in a bocce ball tournament. My fourth year I won best BCB talk (and Megan won best GMB talk). And my fifth year I won the Terry Magnuson Award (which boiled down to being the most try-hard grad student).

To continue echoing the primary theme of these reflections, the best part of each retreat was forming new bonds with coworkers. There’s something distinctly different about spending time with people outside of work. I can’t say anything deeper than “it’s about getting to know one another as people” or something similarly generic. Regardless of exactly why it occurs, the net effect for me—and likely the whole department—was an increase in collaboration opportunities and, subsequently, total productivity.

The Defense

The day after I got back from the retreat, I started prepping for my defense speech. I had two weeks to prepare. As with writing, I’ve given enough talks at this point that I’ve developed a method for practicing, though both presenting and practicing for presentations are things I hope to continue improving. Currently, I primarily brute-force practice. It’s highly effective but perhaps not efficient. Towards the backend of the two weeks, I had to pace myself because I was beginning to lose my voice from talking too much.

I have a single piece of advice for a PhD Defense.

Have fun.

You’ve been working toward this day for the last n years. Enjoy it. Customize the day to fit your desires. If you want to invite everyone who might be interested, you should keep a majority of the presentation super high-level and make a bit of a show of it. That’s what I did. Alternatively, you can invite a handful of people who are going to track what you’re saying no matter how detailed you get, then go home and have a drink. I can’t tell you what to do; that’s your choice. But I can tell you a few things not to do:

• Don’t be ill-prepared. By the time the day arrives, you have hopefully practiced to the point that you’re not concerned about making major mistakes.
• Don’t give the presentation your committee wants you to give. It goes without saying that you should give a presentation that will satisfy your committee, but they’ll likely have lots of opinions about specifics, if asked. Give the style of presentation you want to give.
• Don’t stress. Easy to say, hard to do. But you’re not having fun if you’re stressed. Have fun.

Wrapping Up

I was on campus for a couple weeks after the defense, but there’s remarkably little to say about that time. I reformatted my dissertation after getting review comments, which was a pain. I pulled together my last bits of code and made an exit folder. I got brownies and a really sweet card from my friend Lauren. I had a final lab lunch at Al’s Burger Shack. I said bye to a lot of people. I finished packing the house. I moved on to the next thing.