Doing a bioinformatics PhD: Simple advice on how to plan your PhD and avoid disaster

Simple advice on how to plan your PhD and minimize the risk of disaster. I first talk about when to take courses and which types of courses I recommend, which is courses directly related to your project or transferrable skills such as scientific writing, presentation technique, and project management. I will then explain how to make a project plan that maximizes efficiency and minimizes risk of failure, and the dangers of relying on future data. Afterwards, I will give some advice on how to do the actual project work and how to publish during your PhD.
0:00 Introduction: simple advice for new PhD students and supervisors on how to avoid disaster
0:23 Course plan: systems vary greatly, earlier is better, project relevance, and transferable skills
1:05 Project plan: best part of bioinformatics, multiple projects, risk reduction, stacked risks, dependency minimization, and parallel work
2:06 Data foundation: worst part of bioinformatics, data reliance, data delays, future data, and contingency plan
3:14 Doing the projects: look at the data, automate analyses, use version control, create example data, and focus on the science
4:38 Publishing: writing a review or not, safe first papers, collaborative papers, and the main paper

[@foteiniaktypi3122]

Thank you for yet another great video with very much needed advice 🙂

[@martinhofmann-apitius408]

Really nice, Lars. Will point at this video in my next tutorial for bioinformatics students ...

[@larsjuhljensen]

Thanks so much Martin!

[@glassfabrikat]

Great advice! Thank you! :D

[@larsjuhljensen]

You are most welcome!

[@maasha2001]

My favorite disaster starts like this:"Come and do this project, we are getting this exciting data from . It should be available in a few weeks ..."

[@maasha2001]

And you mention it beautifully a couple of minutes in!

[@larsjuhljensen]

Yes it's a classic!

[@maiaallman4635]

If I want to do research to determine which factors, and to what degree, determine blood glucose levels (against a time scale), does that count as bioinformatics?

[@larsjuhljensen]

It depends on how you plan to do the research more than on the question. If the bulk of the work is to do experiments to answer your research question, then it is not a bioinformatics PhD (even if the analysis part may be bioinformatics). If your work is to analyze data but not to produce data, then it would fit into what I very broadly call bioinformatics here. Medical informatics would probably be the more accurate term, but my advice would be exactly the same :-)

[@anngu2533]

Thank you! I’m interested in doing network analysis on modelling microbiome activities. Does that belong to bioinformatics?

[@larsjuhljensen]

Certainly, but the big question is where the data to create the network would come from?

[@anngu2533]

@@larsjuhljensen I searched for related articles, most of them are based small scale dataset. That’s the problem indeed.