Recently I realized a significant career milestone had passed without my noticing. It has been a few months now since I passed the 10 year mark in my bioinformatics career post-PhD, and discounting work I did prior to it. This number is relatively small compared to how much longer I still need to work, as my financial advisor depressingly reminds me, but struck a chord nonetheless. After some deliberation (and whisky) I decided to write this post. Catharsis by reflection but hopefully to shed some light for newbies to this field. I don’t presume to have all the ‘answers’ and am very aware that what I write below is a combination of my experiences (across academia, biotech and big pharma) and my own nature, but hopefully this will be of interest to some of you. I won’t argue if your observations/thoughts differ from mine (but would love to hear if that’s the case!). I don’t consider my opinions to be facts, and don’t want to impose them on anyone - I hope the same courtesy will be extended to me.
Bifo as a career
Well I’m still here. Bifo has provided me with a career that has had its fair share of ups and downs but has been ultimately rewarding – financially, mentally and scientifically. However it isn’t for everyone, and you won’t know till you try with a decent amount of investment. I have seen a fair few bioinformaticians, for a wide array of reasons, move out of bioinformatics to a range of roles from software/infrastructure development, to data science in other fields and entrepreneurship. I too have made significant forays outside of bioinformatics but have firmly come back to it as I enjoy the science too much. Even if you can’t imagine your entire career in this field, you pick up some very relevant skills as data-driven decision making is an increasingly influential force in the world at large.
The bifurcation of data engineers and data scientists
I was one of many who initially scoffed at the relatively sudden popularity of the term “data scientist” (“We’ve been doing that for years!”) and the near ubiquitous association with AI/ML (which are just a subset of a bioinformatician’s tools). However I have since embraced it and even describe myself as one. As interesting to me has been the rise of the role of the “data engineer”.
Initially I saw this separation of roles as just a logistical necessity in big pharma (where everything is done, literally, at an industrial scale), and I wasn’t a big fan of it – I felt that, on principle, the same bioinformatician should be handling the data from the rawest level, perform all the processing and understand everything done to it, and all the downstream/integrative analyses and interpretation. However I accept that this ideal isn’t feasible any more in most settings. I put this down as much to scale as to the increasing complexity and demands of nearly all steps of data generation, processing and analysis. The complexity and range of data types (this-omics!, that-omics!, multi-omics!) is ever increasing, as is the range of ways you can process them (how many aligners are there?) and analyse them (stats, systems/networks, AI/ML), as is the infrastructure and tools needed for these purposes (R!, Python!, Cloud computing!, NextFlow! Docker!). Bioinformaticians are interdisciplinary scientists already (more on that below) but to be able to keep up and be an expert in all parts of everything is a hell of a job – or maybe I’m getting old. However as roles start to diverge, it becomes ever more important that the data scientists and engineers are communicating often and deeply.
Impostor Syndrome X 10
Given how even seasoned professionals within relatively narrow focus/ specialized fields can suffer from, at times, mentally crippling levels of impostor syndrome it is natural for interdisciplinary scientists such as bioinformaticians to suffer this even more acutely. It is unfair of anyone (particularly yourself) to expect you to have the level of statistical insight of a statistician, the coding/infrastructural nous of software developers/engineers or the biological knowledge of biologists. We all have varying levels of these, with very few being “experts” across the board. My advice would be to embrace our “inadequacies”, confidently explain this, and decide whether there any areas that actually need addressing/development for the purposes of your specific role.
At the start of my career I used to beat myself up about these continuously. A watershed moment for me came in conversation with a highly experienced, knowledgeable, and universally respected expert in his field, and he said he always felt he was in “fake it till you make it” mode. I suppose the inverse of the Dunning-Kruger effect is that competent people never feel they have “made it”.
Beware the “embedded bioinformatician” role
This is speaking as someone currently in this sort of role myself. Why I bring this is up is because while I know many people in this position who are quite content (myself included, at least for now!), I would estimate that the majority of bioinformaticians whom I have spoken to who are unhappy/unsatisfied with their positions are in embedded roles. So many times have I heard the same heartbreaking story of young, highly intelligent, diligent bioinformaticians slowly have their confidence, self-worth and drive whittled away due to similar themes: being treated as a glorified technician, having their opinions ignored/not valued, being held in lower esteem to “biologists”/”real” scientists, their needs – tools, development – being a lower priority to others.
I feel much of my own satisfaction with my current role is that, aside from having skills and experience that have given me a modicum of self-worth, I have having come to terms with the fact that this is how things are and that it is up to me to do something about it, if I can. I have tried to be confident in asserting my opinions, open about my concerns, and assertive about my needs with the leadership. I have no doubt this would have been much harder in earlier stages of my career. While I don’t want to put off newbies from embedded roles, I would only advise that you are aware of the issues, be open about them with the leadership (bring them up early, even at interview – I did!), and make decisions based on how the leadership reacts. If the leadership is on-side (or at the very least pretend to be), other things can fall into place. Ultimately your work and how you present yourself is what will buy you respect.
Thanks to the community
I can’t overstate the importance and influence of platforms like Biostars, the Bioconductor forum, SeqAnswers, Stack Exchange and many others, have had on both my day-to-day work and my career in general. I would like to end with a massive thank you to the community of bioinformaticians online who are helping each other out. You are all stars.