How to transition from pure biology PhD to bioinformatics/biostatistics?

Question

I've been working as a PhD student in biology and currently in my last year. However I realised just a year into my PhD that I hated doing experiments, I really hated it. I think one of the reasons may be that in the beginning, most of my experiments did not work. Also, I couldn't justify spending years of my life working on a tiny protein that no one cared about, all in the name of basic science. I initially wanted to quit but it was a difficult decision because I am an overseas student and may have had to immediately return home if I did.

During this period, I did find that I enjoyed working on statistics and taught myself to code a bit, mostly at night. I'm currently taking online coding courses and I really enjoy using computational methods within my field especially in statistics. My main question is: is it realistic to switch to a career in bioinformatics and statistics, and how can I go about this immediately after my PhD?

Answer 1

As a PhD student in Math with Computer Science background who is working in Bioinformatics exclusively: it is definitely possible for you!

In the Computational Biology Institute at The George Washington University where I work as GRA we have PhD students, Post-docs as well as Faculty from different backgrounds (including pure biology) working purely in bioinformatics area.

I do believe you don't have to switch as much, as you think. It only takes a group that is more focused on the "informatics" part more, than on biology. But still caring about Biology to change the main daily routine of yours. And your background will be needed, as bioinformatics can not exists without biological reasoning and biological guidance / insights.

If I were you, my first step would be to find what is the most "informatics" related problem close to your current research. It might be not your group, who is doing it, but still. Educate yourself a bit about such research by reading related articles. And after that it is as simple as writing a couple of emails to PIs of respective studies and telling them, that you are very interested in collaboration on their topic. This might yield a joint publication or even a post-doc position offer from respective lad.

I wouldn't count much on pure "career" in statistics, as it usually requires more math background. But biologists in bionformatics, who at least knows how to do statistics are very hard to find. So the more you know from the programming / statistics side, the better you are off go to more informatics groups.

Answer 2

No! Don't run into anything. Think what is the real problem. Is it that you don't like dealing with uncertainty?

"I hated doing experiments"

What makes you think that you will not do experiments in "bioinformatics and statistics"? An Experiment does not have to be something done in test tube. People working in bioinformatics and statistics are also doing experiment all the time, they are experimenting with there code and methods. And yes most of the time things do not work. After all it is "Research".

OR is it

I couldn't justify spending years of my life working on a tiny protein that no one cared about

The bad news is that it is true in other fields also. There are plenty of bioinformatics/statistics methods/papers that no-one cared about (My gut-feeling: such papers are more common in bioinformatics/statistics). No-one can predict what with absolute certainty what what will work and about what people will care.

Bottom line, experiments not working and no-one cares about my research are common in science. If you think that this is the main problem that what is the guaranty that you will not face them in other field? In that case you should also consider option of moving out of academia.

In theory it is possible to switch to subject and it might be easier because fields are linked. But you should not forget the bigger picture and reality. You don't like your current topic, that indicates that might took a ill-informed decision in past. Don't repeat it again. Before taking any decision, talk to your Professor or someone who is experienced and know you well (about your research work). And most importantly talk to a person who is doing bioinformatics/statistics. Someone whose job you want ideally in future. Ask them about the day to day difficulties and challenges. Put yourself in their shoe and then decide.

Note 1: If you want to work as a bioinformatics/statistics technical support then it might be little different (less experiments). But as far I have know people at technical support side of the science are also doing "experiments".

Note 2: As others mentioned, people coming from different field to bioinformatics/statistics, it is true BUT current situation is more different then 10 years ago. As bioinformatics as a field has evolved and matured more and more jobs required more specific knowledge and skills. Although it is not impossible to gain skills but my point is it is much difficult now a days to change to bioinformatics as it was 10-15 years ago when there were very less bioinformatics specific students/phds.

Answer 3

Yes you can! As you are motivated enough to learn stats or computer programming, without spoiling your last year, I suggest you to try to publish a work (w/ collaboration with someone in maths, data science, CS), at least in a conference, with a twist toward bioinfo or biostats. This will show the bend in your carreer to future employees.

Answer 4

Its definitely possible, I was in the same position as you, and I made the same change myself, although I did this after a wet-lab postdoc. I am now a bioinformatics PI.

Personally, I was lucky enough to land a position with a 3 year program specifically designed to take wet-lab trained biologists and retrain them in bioinformatics, but its definitely not the only way to do it.

The absolute best way to do it would be to undertake a small computational/statistical project relevant to your current work. It will go much better and you are more likely to succeed if you find yourself a friendly bioinformatician willing to help you with the project, but don't approach them cold and just say "I want to learn to do bioinformatics, please give me a project". Rather say "I've got this idea, and I wonder if you'd be able to help me realize it".

Even in the absence of this, you might find that you will be able to land a bioinformatics postdoc. There is a massive shortage of trained bioinformaticians, and if you bring some enthusiasm, you might be able to land a position. But two words of warning:

1. Made sure you will have the supervision of someone who knows what they are doing - don't volunteer to be a pet bioinformatician for someone who can't supervise you without any other support.

2. Think what you bring to the relationship - perhaps the position is 50:50 wet-dry and they are looking for someone with both skills - so you may be required to do some wet lab for the opportunity to learn the dry lab. Alternatively you may being specific biological knowledge that a supervisor is willing to exchange in return for a training in computational biology.

As has been pointed out above, probably you will always be in a particular part of bioinformatics - the part that applies computational tools to biological problems, rather than the part that applies computer science to create tools.

Answer 5

Apply for an M.S. in Statistics or Computer Science, this can help you get a good background. Then apply for Post-Doc positions. An even better idea is to get your M.S. in Data Science, and then find a post doc in something of that sort. Data Science is a highly lucrative field, and if you have an interest in it, Why not?