So You Finished Your PhD in Computational Biology. What’s Available in Industry?
For a lot of scientists, it is common to finish a whole PhD in their field without ever talking to someone in industry. Soon-to-be graduates will sometimes reach out to a friend in industry to ask about jobs. However, industry offers a wider variety of jobs than the well-defined academic track, so it’s difficult to get a well-rounded view from just one perspective. Some members of the Bits in Bio community decided to put together a longer answer to provide another perspective on being a computational biologist in industry. You should also read Elliot Hershberg’s post on his excellent blog about the intersection of postdocs and industry.
Let us start by saying: it’s OK to leave academia. There is nothing sacred about jobs in academia, and you don’t need anyone’s permission to choose the career path that’s the best fit for you. There are many great things about academic science, but there are drawbacks as well.
One major difference that is important to highlight is that academic pay is usually much lower than industry pay. Here are the Glassdoor listings of salary information for bioinformatics scientists in Boston. It’s noisy data, but $115K as the lower end of the range sounds about right. When you have a few years experience, or you start getting out into Silicon Valley, or factoring in stock options, the top of the range can be anything. There are also plenty of good resources on negotiating salaries and what are stock options if you poke around the internet.
However, it is also worth noting that there has been a tiny, little 30% drop in the biotech market over the last year (see the $XBI). Some very good people have gotten laid off. But hiring had been difficult prior to the drop (with jobs open for months), so it seems unlikely that computational biologists will stay out of work for very long.
This post will walk through some tradeoffs to think about while finding a job. Keep in mind that neither of these options is better than the other, nor are they exclusive. This post is meant to help you find your own path. Here’s what we’re going to cover:
1) Start your own thing or Work for someone else
Starting your own thing is not for the faint-of-heart.
However, many people end their PhD programs with a project or insight that is better taken forward in an industrial setting than an academic one. You may have a technique or a therapeutic target that can benefit the world in a way that will never happen if it stays in academia. And there is no one more qualified than you to do it.
Going down this road is hard work and may not be what you want to do, but this is still one of the first questions you should ask yourself. Your university should be a good source of information about this. Pillar VC has some great resources here as well. If you decide to move in this direction, Nucleate is set up to help you along this road.
2) Individual Contributor or Manager
Individual contributors don’t have anyone report to them. Managers do.
Most people start off as individual contributors. If you really like doing technical work and don’t enjoy managing people, then this is a good role for you. Senior people who are still individual contributors are usually there because that’s where they want to be. Being an individual contributor is a viable and valuable career path!
This is quite a different mindset than in academia where success is often tied to having a huge group with as many people as possible. In contrast to how most academic labs are structured, computational teams with 20 year industry veterans are often managed by much younger colleagues because the veterans don’t want to be managers.
So yes, you can be an individual contributor your whole life and just do science.
You can also be a manager, but it typically takes a few years to grow into that role unless you have a specific skill set or start your own thing. Managerial roles are nice if you enjoy solving people problems and mentoring junior scientists. Project management is a big part of the job. Some people love that. Some find that tolerable at best. Managers focus on building a great team and their success is no longer tied to their individual work, but rather their team’s.
Everyone’s career path is unique, but it’s typically easier to grow into a managerial role at a smaller, earlier stage company. If this is what you want, it is useful to ask questions about how many managers have come from internal promotions during the interview process.
Some places are very hierarchical and some are very flat. Titles can vary widely across companies, so make sure to ask specifically how big the groups are, how many managers there are, and what sort of promotion process and management training you can expect.
Being able to do both is a certain kind of sweet spot. If you go too far down the managerial path you may find it difficult to transition to a new venture where you have to be an individual contributor for a while until you build out a team.
3) Digital Product or Biological Product
What kind of thing do you want to work on?
There are two ways to integrate bioinformatics into a biotechnology company. Some companies are digital-first. In this model, a company’s main asset is its computational technology and biological experiments are either nonexistent (such as in a SaaS company or bioinformatics consultancy) or play a supporting role to the digital platform. These companies are structured much more like tech companies, with the plurality of staff being computational.
In other companies, the roles are reversed. The main product is biological and the digital platform exists to enhance and improve upon the wet-lab product. The plurality of staff are wet lab scientists, with computational teams making up a much smaller percentage of the overall staff. This is how most biotech companies have traditionally operated.
For example, a digital-first company may have a machine learning model as its primary product. The company has a wet lab, but its main function is to generate data to feed into the model. In a biology-first company, the main product would be biological (e.g. a therapeutic) and the machine learning model is used to understand the experimental data and improve upon the wet lab results.
Some companies do both! Companies that started out selling software or hardware have started producing drugs, and vice versa. There is a lot of interesting work happening now across the whole spectrum of digital to biological — there’s no right answer here.
4) Tool Builder or Tool User
Tool builders make new bioinformatics tools. Tool users select, implement, and parameterize tools made by other people.
Tool builders get all the glory in academia, but there is a huge need for tool users within industry. In academia, there is a lot of work associated with marginal improvements in algorithmic performance. In industry, the ability to evaluate whether an existing tool is “good enough” and then put it into practice is in high demand.
People who trained as a tool builder can usually be both a tool user and a tool builder in industry.
In industry it’s much more common to only build tools as a last resort.
5) Software Engineering Best Practices or Take on Technical Debt
Computational biology departments can vary a lot depending on: who leads the department, who is working there, and how many members of the team there are. A large department sharing a codebase should have a lot of software engineering tools and robust best practices.
These groups usually implement processes and automated systems to facilitate working on complex software projects. This includes git for version control, processes for approving changes to code, writing automated tests, and adding checks to ensure the code meets coding standards. This infrastructure is much more common in the tech world than academia.
Other companies don’t look like this at all. In many places, people work as individuals on their own pipelines and notebooks. This is often more common in earlier stage companies where the speed of getting things done is more important than getting things into a shape that would impress your computer science professor. This approach can serve early stage organizations well because it helps you get to the next step in your growth more quickly. Also, when things are rapidly changing, a lot of technical debt that you take on will never have to be paid. For example, if you write sloppy code to fix an artifact in the data, you don’t have to clean up that code if the wet lab cleans up the artifact first.
Many successful companies have been built on the backs of shoddy Unix scripts, and it’s normal (and desirable!) to take on some technical debt during the early stages of a company. But as organizations continue to grow in this way, they start to spend more time fixing bugs than developing or using tools. Once sloppy development begins to cost more time than it saves, companies will start to invest more in computational infrastructure and best practices.
This may seem like a trivial distinction but it is really, really important to have a good fit on this one. There is a substantial and frustrating learning curve to going into a mature development environment, particularly if the subtleties of git-rebasing your code is not anywhere close to where your interests lie. Good people have quit good jobs over being bad fits here.
That being said, every company that survives long enough to build a messy code base eventually transitions into a place which introduces more robust software systems and best practices.
6) Early Stage or Late Stage
As you may have gathered from the last point, the computational biology departments of early and late stage companies look very different.
The stages of a company roughly break down along the lines of where their funding is:
Pre-seed funding: The founders have an idea and a slide deck but little to no money yet. If you are joining here you are probably a founder or know the founders and aren’t reading this blog.
Seed stage: The company is based on an idea that someone thought was good enough to fund with a few million dollars. The company will be laser-focused on providing proof points that their technology works as they described. The focus is to not run out of money before being able to raise a Series A.
Series A: This is typically the first big round of funding, usually from a consortium of private investors called venture capitalists (VCs). These rounds are usually in the tens of millions but can be in the hundreds of millions of dollars. This money is used to push the idea from something that works in the lab to a more sustainable company.
Series B - Series ???? : These are additional funding rounds to supplement the Series A. There’s usually a series B and sometimes more private funding such as a Series C,D,E…
Exit: This is an event that happens when the previous investors get their money back. This can be by taking the company public (IPO) so that additional funding rounds come from selling more stock. It can also be through acquisition where a bigger company buys a smaller company. This is where the employees can sometimes turn their equity into cash.
Public: Going public is often seen as the final state for a company. This doesn’t mean public companies necessarily have a lot of money or are particularly stable. We would include both big pharmas and biotechs in here. Big pharma companies are unlikely to collapse altogether, but go through ups and downs depending on the market and their clinical pipeline. Public biotechs, on the other hand, do sometimes collapse altogether. An all to common story is for a biotech to get to an IPO and then live the rest of its life on a slow and painful declines until it gets delisted from the NASDAQ.
Joining a startup (stages 1-4) has more risk but usually promise a big stock option payoff if they’re successful. Startups also provide the opportunity to work on novel technology, solving a problem that hasn’t been solved before. Thinking through the risk-reward tradeoff is important, but don’t forget to factor in your intellectual interest when making a decision.
More established companies may pay you in salary plus stock. This is typically a less risky option, but even big companies can lose a lot of value in a market correction.
Whatever path you take, don’t forget to negotiate!
7) Big company or Little company
Bigger companies are almost always later stage companies. No one hires 1000 people on their first day. The size of the company has a number of ramifications on both your technical and non-technical work. Earlier, we discussed some of the technical changes that happen as your company matures its way of developing software.
There are also a number of changes that happen to your non-technical work life. Big companies typically have better operational infrastructure and a certain level of institutional professionalism. This can manifest as having a defined onboarding process, an HR department, and processes for suggesting new projects to work on. A startup often doesn’t have the time or expertise to put this infrastructure into place and is much more reliant on individuals making the best decisions they can without the support of institutionalized processes.
Size and stability might be of paramount importance if you are reliant on a visa or other immigration-based work authorization. Smaller companies are less likely to sponsor a visa and are unlikely to be able to provide much guidance in navigating the immigration process. We recommend talking to people who have gone through this.
8) In person or Remote
Industry labs, in general, did not close during the pandemic. The computational biologists were, rightly, chucked out of the offices so they wouldn’t be breathing on the wet lab staff who had no choice but to come in.
As a result, many computational biology positions are now either full remote or hybrid.
People either love or hate working remotely. Some computational teams are in the office every day now. Other computational biologists will never see the inside of an office again.
By this point, you probably already know what side of the divide you are on. Both job types still exist in the field, and you should just ask how it’s handled. This, again, is something you might consider a deal-breaker if it’s a bad fit.
9) Purely computational or Hybrid wet lab
Most computational biology jobs just involve you writing code. However, some people have jobs where they do both computational biology and wet lab work. We know of someone who runs the whole NGS operation at her company. She does all the wet lab protocols and then she analyzes the data. These jobs are less common but they do exist.
You should assume that your job will be purely computational unless otherwise specified.
10) Toxic or Nice
You want to go to a nice company. Finding a nice company can be tricky.
The first and foremost is that individual experiences at companies differ a lot. At a big company, your experience will be largely dictated by your department, and maybe your individual manager, and less by the company as a whole. For example, the wet lab might be miserable while everyone in comp bio is happy as clams. Also, the toxic aspects of a company (and every company has some) affect people differently. Maybe the culture is super misogynistic but your coworker doesn’t notice it because he is a dude (work on that, bro).
Here, Glassdoor is your friend. People usually won’t leave a bad review unless they really hated a place. Look at what they say, when they were written, what roles people were in, and how many reviews say the same thing.
A small startup won’t have any reviews, so you have to figure out what is going on during the interview process. It is nice if you can talk to employees when their manager isn’t there. Once, someone we knew was left alone in a room with a guy who told them not to work there because the company didn’t have any money. That company closed a few months later. Trust your gut during the interview process — if the vibe feels off, you’re probably onto something.
If you do your best but still find yourself in a toxic situation (it happens):
1) Don’t contribute to it. At all. If you have a PhD you are a senior member of staff and your actions have an oversized effect on junior members of staff. Act like it.
2) Leave. Don’t wait to get your bonus. Don’t wait to get your stock vested. Concentrate your energy on finding a new job and leave. You can take it off your resume if it’s very short (<3 months). If you have a lot of short stints (<1 year) you have to start explaining that, but if it’s only one, and especially your first one out of grad school, no one cares.
One of the benefits of working in industry is that it’s not a monolith. You only need 3 references to say good things about you, so you don’t need to put up with being treated poorly or watching other people being treated poorly.
11) Computational biology or Bioinformatics
They are the same thing.
We hope this post has given you some things to think about. You should make a list of your preferences in your head, but don’t close yourself off to opportunities if something doesn’t exactly match your list.
An additional benefit to working in industry is that you can switch around every few years without suffering any reputational cost.
You should definitely take advantage of the amazing Bits in Bio community to ask more questions and meet others who have made this transition.