Michael Barton, PhD

Bioinformatics Systems Analyst
Joint Genome Institute

Website: michaelbarton.me.uk
Blog: Bioinformatics Zen


Unfortunately there was no literal moment of inspiration where I decided that I wanted to do computational biology. I had done very poorly in my undergraduate degree in molecular biology and so I found it very difficult to get an entry level job in the field afterwards. I remember sending out many resumes and getting nowhere. Ultimately I had to get any job that would pay, so after my degree I worked various roles including a door-to-door salesman, answering telephones in a call centre, and teaching English as a second language.

After two years doing these jobs, I wanted to do something that gave me more fulfilment. I remember talking to my sister, who works in microbiology, and she suggested that I try doing bioinformatics as it was a new and interesting field. I thought it did sound interesting and so I applied for a couple of masters degree courses around the UK. I was accepted to the University of Newcastle. I really enjoyed this course and this is how I was able to start pursing a career in bioinformatics.

  • Dedication. This is the one thing that I most recommend to anyone pursing a scientific career. Science is very exciting but sometimes it can be extremely boring. During these boring times is when you just have to put the hours in, no matter how dull it is. Things like revising a manuscript or updating figures for what feels like the hundredth are what's required to get something finished and completed. Once something is published you can look back and be proud.
  • Continuing education. I don't think education finishes when you complete your university degree. I think you have to keep studying, this can be keeping up with the literature, learning a new programming language, or teaching yourself linear algebra. The most exciting and best people to work with are those who are hungry to learn new things.
  • Aim to excel at what you do. I think when you start something you have to aim to do your absolute very best at it, otherwise why do it? Learn to say no to projects you think you might do half-heartily and commit 100% to the ones you say yes to. Always try to pay attention to the details as well, this can turn a good project into a great one. If you always do your best, other people will eventually start to notice, or if they don't go somewhere else that rewards your hard work.
  • Surround yourself with very smart and hardworking people. There's only so much you can motivate yourself sometimes. If you work somewhere where everyone else is working hard and doing great work, you'll want to do the same. You'll start pushing yourself to be even better.
  • Take time to relax. I think it's very important to have balance, you can't spend all your time working or you'll eventually hate it. I make an effort to not to do work or check emails on the weekend or in the evening. I try to work as hard as I can during the day, and then relax otherwise. I don't go home and watch TV though, instead doing something active helps me relax. Right now I've been learning to surf, and a few hours each day on the weekend helps me unwind so that I'm fresh to start again on Monday.

I learnt a great deal during my master's degree at the University of Newcastle. I learnt Java and R, as well as many of the basics of bioinformatics. Since then I have mostly been teaching myself through books and Coursea online courses. I am always interested in new programming languages as so much time in bioinformatics is spend writing code. Learning functional programming with Haskell was a really eye-opening experience that showed me how much better writing software can be. I like learning math and statistics and this usually from books and videos. A good grasp on linear algebra and statistical methods can really help you tackle difficult problems and get good answers.

I work at the JGI so most of what I do focus on sequencing and genome assembly. We look at what gives a good assembly and how various metrics in the sequence data relate to this. One current project - nucleotid.es, aims to provide objective benchmarks for genome assemblers.