Sunday, 27 May 2018

Leaving academia: failure or chance? A personal story

This blog post will be a bit more personal, as I will talk about my career, the paths I took and the mistakes I made. This might be useful for other scientists in the same situation, despite every case is different.
[I am voluntarily omitting any names of my collaborators, supervisors, teachers, but I am grateful for all of them for their advices and support]

Undergraduate studies

I was born in a town close to Lausanne, Switzerland. So, naturally, I did my undergraduate studies in biology at the University of Lausanne. Nothing really particular to mention there. The first two years (“propedeutics”) were generic (botanic, zoology, molecular biology) and the next two years were more specialised (biochemistry, immunology). Anecdote: I had the chance to follow the lecture on biophysics from Jacques Dubochet, the recent Nobel prize. He is the kind of professor that made a course alive. One thing he mentioned recently is that he wouldn’t have been able to do a similar academic career today due the intense competition for everything.

During my third year, I did a internship in immunology. It was a good experience, but I destroyed quite a bunch of glass tubes and other tools in the lab. This was a clue that I was not made for any experimental labs, and that the only mice I would then torture would be from Logitech or Apple. During my fourth year, I had the great opportunity to do my master project within the Swiss Institute of Bioinformatics (SIB). I was working with sequences alignments and homology modelling of 3D protein structures. It was a blast and the moment where I knew how to orient my career.

One drawback for studying biology is that most courses are mostly based on strict learning, and not much about problems-thinking. We only have to learn what the professor taught and write it down at the exam. In the end, very little thinking involving computation or equation solving. As I have a very bad memory, I was struggling with some courses like biochemistry or zoology, while I excelled in other courses like physics or java programming (because we could have notes during exams). Another thing I lacked during my undergraduate was a modern course in statistics, and practicals in R. A shame that the current professor of statistics arrived only a couple of years later. I know his lectures are good, because I was involved as assistant for the practical during my PhD studies.


PhD study: 2005-2010

I started my PhD in 2005, still in Lausanne, with a four-year contract. It seems rather long compared to international standards (maybe not US), but I think it is a good length, as I was quite productive in my last year. I worked on computational biology, especially in molecular evolution and phylogenetics, at the Department of Ecology and Evolution. It was a good period of my life. We frequently read that we could be lost/depressed/down during the PhD, but I never had this feeling. It could be due to a good mixture of supervision/environment. There was definitively something special in the department daily life [1]. Lot of social and professional interactions between members, whether you are PhD student, postdocs or group leaders, and a rather flat structure.

A good advice I received from my supervisor was to not wait the end of the PhD to search for the next step, but start earlier. So in 2009, I applied to three fellowships to go to University College London: Swiss National Science Foundation (SNSF), Marie-Curie and EMBO Long-Term. I was rejected to all three during the first round. Some reviewers said my project was not ambitious enough, other said I haven’t had enough papers in good journals (the same who think that bioinformaticians just need 2-3 clicks to process data and publish in Nature...). Luckily I managed to obtain a short-term contract for 2010, right after my PhD. I then re-applied in 2010, and finally obtained the SNSF fellowship. While I applied for the maximum number of years, I was awarded only 1 year of fellowship, due to the University of Lausanne policy. One year is better than nothing, but still, there is not much we could in one year in my opinion.

1st Postdoc 2010-2013: University College London

In August 2010, I flew to London, to join the Department of Structural and Molecular Biology. I worked on a molecular evolution project, but with a deeper focus on 3D protein structures. I stayed three years and it was very good in term of academic freedom, as I came with my own funding. So my time was split between developing my research projects, trying new things (that didn’t work, eh this is science) and participating in the lab’s databases development.

One drawback during my first year is that I had only one year of fellowship, so I wasn’t sure what would be my life after August 2011. This provided a little sense of uncertainty during the first year, and kind of emergency to have to some results at least published in a year. No need to say that in these conditions, I wasn’t planning to do something very ambitious and very long. I also spent some time writing my application for the Advanced SNSF researcher, which I got it and was awarded for two years this time (the maximum allowed by the SNSF). In 2013, I applied to come to Switzerland, but this didn’t work. Luckily, I got another postdoctoral contract at EBI.

2nd Postdoc 2013-2017: European Bioinformatics Institute (EBI-EMBL)

In 2013, I joined EBI-EMBL, in Hinxton, Cambridgeshire, UK. I stayed there for nearly four years, until April 2017. I worked on a phosphoproteomics project. While these years were great at the scientific level and with colleagues, they were quite tough as I started to apply to other, more permanent position in academia (the doom of fixed-term contract for postdoc). By more permanent, I mean tenure-track professor, lecturer, principal investigator or group leader, depending of the institution or funding agency. I applied to almost 50 positions, with barely no success. I was invited for interview to a few of them, but competition was fierce. One of my numerous problems was that my proposed project was not interesting enough. Another problem is I didn’t put an experimental component in my proposals, which would allowed me to target high-impact journals. Newly generated data can be re-used by other and increase the paper citations. I also started to apply to industry position, lightly at the beginning, but more intense as the end of my contract arrived, along with multiple rejections from applications in academia. Luckily, I managed to get a position in industry before the end of my contract. I will write about that in a future post.



Some statistics and various comments about my applications:

I grouped my applications in three sectors:
  1. University, which speaks for itself.
  2. Academia = other non-profit institutes such as EMBL, Pasteur, Crick. They tend to have better funding and higher salaries than local universities.
  3. Industry = pharma and biotech companies.


Timeline of my applications by month and by sector (2013-2017):


Comments:

My first attempt was in February 2013 for an SNSF Ambizione fellowship. It was just before starting at EMBL-EBI. The success rate was 40% when I decided to leave Switzerland in 2010, but suddenly dropped to 24% in 2013. Depending of the department policy, you can either be a group leader or a “super-super-super post-doc” (not my words) attached to a lab. I was invited for the interview at the SNSF headquarters in Berne. I apparently did a good presentation, but I failed the questions part. I was definitively not well prepared for that. I reapplied for the same scheme in 2014, but not even been invited for the interview this time, and that was the last possible time (maximum five years after PhD). Then I applied for a SNSF professorship in 2015 and 2016. This grant is really good and provides all the money you need to run a small lab (PI +PhD students + postdoc) for 4 years, extensible for 2 more years. Again, other failures, because my level of publications was good, but not excellent (aka not published in Cell , Nature, Science [CNS]). Then, since Summer 2015, I applied to other positions at various places, which I will describe below. 

Number of applications by country and by sector:

 


Comments:


  • As a Swiss national living in UK with family, there is no big surprise that I mostly applied in both countries. 
  • Switzerland is extremely competitive for academia. Being a national doesn’t give you any privilege, only scientific record matters. Salary is huge (but living there can be very expensive too, people from abroad don’t necessarily see that), so many people from around the world apply there. For each position I applied, there was between 100 to 250 other applicants. The problem now in Switzerland is they put the tenure-track system. It is not a problem per se, but there is the risk  that after 4-6 years, you cannot stay there. Which means the need to find another job at 40-42 years old, which might not be so easy. And even worse for the SNSF professorship, as the host university doesn’t necessarily have a backup plan to keep the professor after his/her grant expires. But this non-tenure grant system is not restricted to Switzerland and seems to be more and more common around the world. 
  • UK can also be very competitive, depending on the place where you apply. I was invited for an interview at University of Leeds. It is was a very positive experience, I met a lot of interesting people, but in the end I wasn’t selected for their scheme. The main reason was that my research topic was not cutting-edge, so it might difficult to apply to grants. I also applied to some pharma/biotech companies around the London/Oxbridge triangle. Surprisingly, industry is less competitive than academia, as there are much fewer applicants per position (at least in bioinformatics). I always rapidly got an answer and most of the time a phone call. However, I had two handicaps there: 1) I haven’t had previous experience in industry (except a six-month training in Nestlé before my PhD, so looooong time ago) and 2) I didn’t have a sufficient exposure in genomic sequence analysis (NGS, RNAseq, WGS, GWAS), which was most of the time a required skill. Eventually, I did end up working in a biotech company in London, which will be the subject of a future blog post.
  • In France, I only applied to the Pasteur Institute. The first time I applied, I was not invited. I re-applied six months later, and then I was invited for an interview. I was told they spotted my application in the previous round, so they decided to give me a chance this time. But in the end, they funded people with research topics more in line with the institute (and again, I kind of screwed the question/answer part during interview).
  • US: I applied to some places in the US, but was never invited. I guess it is even harder to go there from Europe if you don’t have any network.

[Introspection] Some reasons why I failed in academia:


  • Not working on cutting-edge topics (next-generation sequencing, gwas, single-cell, you name it). While phosphoproteomics is quite fancy these days, it is still less marketable than genomics. I also haven’t put any experimental component in my grant projects, and new data is definitely needed when you want to publish in CNS journals.
  • Didn’t race for the CNS papers early enough in my career. I published my first postdoctoral work in PNAS, and some people call this journal Paper Not Accepted in Science (but no mistakes, I am still proud to have published there, it was the first journal we targeted and the perfect fit for it).
  • During my two postdocs, I should have applied to other fellowships in UK. Getting money from grants increase the success of the next applications.
  • Introverted character. In science, it is quite common to meet introverted people. I made some online test, and without surprise, I am type INT-J with a strong bias towards introversion. Basically, I don’t like to express myself in public, whether it is group meeting or to talk in conference. I am also not particularly keen of walking in large conference halls and trying to talk to people I have never met before. Which could be a problem when you are trying to build your network (so I compensate by communicating on internet).
  • Another problem is I am slightly more interested on the technical aspects (low level) of a project, rather on the general aspect, or big picture (high level). For example, when I review a paper, I will try to replicate the result as much as I can, until I am satisfied. It is useful as a highly skilled engineer, but less as a visionary group leader or manager. I recently had this discussion with a former colleague, and the conclusion is we just need to accept who we are and do what we are good for. And this is a problem in academia, most permanent position are for group leader, who needs managerial skills. Staying in academia as highly skilled professional is more difficult, as there are barely no permanent position as super-postdoc, or you are dependant on the lab budget, which can switch rapidly.

Some random advices (for any level, undergraduates, graduates, postdocs):

  • If you love sciences and have some skills, go for undergraduate studies. And if after that, you think you would do science as your main job, go for a PhD. It is the highest diploma you can get in academia, and many companies in industry are using it as a threshold when doing recruitment for researchers. It also helps to climb the career ladder.
  • But at the same time, you must be aware there was a boom of PhD in sciences, and the academic system cannot sustain the increased demand for jobs. You need to be aware of this at the earliest in your career. You can be among the top scientist of your classes, you can be clever and you can work hard, this might not be enough. At this point, the number of applications per position means there is a stochastic process, where randomness occurs and there is nothing to do against. Just try the best and not be disillusioned if you don’t succeed.
  • To succeed in academia, it is sad, but only publications in CNS journals matters (Cell, Nature, Science). Publications in other top journals (Genome Research, PNAS, Mol Sys Biol, PLoS Biology) are very good, but there will always be someone else coming with such publications in CNS journals.
  • Avoid writing book chapters. I consider it as a loss of time and energy. First, you are not paid for that (but the publisher will make some money). Then, if it is an original study or review: wrap it up to publish it in a peer-review scientific journal (with open access option). Finally, if this is something more technical like a tutorial, just publish it on your own blog, free and accessible for everyone.
  • Try to learn methods that are useful in industry, in case you want to shift at some stage of your career.
  • One year before the end of your current contract (PhD, postdocs), start looking for the next step in your career. It takes time to find a host, to apply for grant, then to re-apply in case of rejection.
  • If you want to do postdoc, I recommand to apply for fellowships (Marie-Curie, EMBO, SNSF, etc...). They can offer a greater freedom and you could more easily find a lab, as you are going with your own funding.
  • If you are hesitant about industry or academia, don’t waste time on a postdoc and start applying to industry position.
  • If you didn’t succeed to publish your postdoc project in Cell/Nature/Science, or at least one of the top journals in your field, start reconsidering your career in academia.
  • Doing a second postdoc means you didn’t succeed to go up to the next level on the academic scale during your first postdoc. At this stage, you should already be on lecturer or professor tenure-track position, or at least on an independent fellowship leading to lecturer/professor position. So it is really time to move out academia and search for another career.

Conclusion:

Despite the story I wrote seems rather bleak, it doesn’t mean failing in academia is the end of everything. I started working in a top-notch company exactly one year ago, and I didn’t regret it. This will be the story for another blog post.

Notes:

(1) Our third area of specialisation consists in having weekly parties and numerous other social events. We realised that mutualism and cooperation lie at the heart of the ecological success of many organisms on earth and therefore make use of this important observation to enhance productivity and happiness. Collaborations between the different research groups is a Department priority, with two weekly seminars and numerous journal clubs taking place. All the resources and infrastructures are shared between research groups, providing a stimulating and nice atmosphere to students and all other members of the Department. https://www.unil.ch/dee/home/menuinst/presentation/introduction-from-the-director.html

Further readings:


5 comments:

  1. Thanks for sharing your experience Romain. Great post as always.

    ReplyDelete
  2. Thanks a lot for sharing your experience, Romain! As someone who recently was invited to a Posdoc interview at a US university, but screwed the questions/answer part, I can relate completely!
    All the best!

    ReplyDelete
  3. Wow that was very insightful Romain and quite the journey! I am about halfway in my PhD career but I am hoping to land industry positions because I like stability more than the competing, stressful environments that I come across in academia all too often.

    ReplyDelete
  4. These are very interesting remarks. I got to know several people that stayed too long in junior PI positions and postdoc positions and then after the age of 45 were let go with no place in academia and so far, for none of them, a place in industry. That is my biggest motivation to leave academia after my first postdoc.

    ReplyDelete