Roberto Kolter set up his microbiology laboratory at Harvard Medical School in Boston, Massachusetts, in 1983. Postdocs worldwide hope to join his lab because of his career-targeted training philosophy, but with rare exceptions, he brings in only those who already have a fellowship.

Why do you accept postdocs only if they have their own funding?

I focus on those whom I believe have a fantastic chance of getting their own funding as a principal investigator. I think it's unfair for me to interview those who have very little chance of getting their own funding, considering how competitive the academic job market is and how important it is to show independence.

Credit: Einat Segev

What does your laboratory focus on?

I let the postdocs explore what they want to explore, as long as it is within the sphere of my interest. I've worked on starvation physiology, biofilms, signalling, experimental evolution, antibiotics and many other subjects.

Describe your training philosophy.

I train people to go on into academia, industry the corporate world or whatever they want to go into. We need to give them the experience that they require, including learning how to teach and learning how to manage. Postdocs are not just there to come to the lab so that principal investigators can get their next grant.

What stands out when you look at applications?

I have learned that networking works very, very well. If I know who trained that individual, and I know and respect them, then I'll know a lot about how this postdoc will work in the lab. But that does not mean that if I don't know the mentor I will close the door to the postdoc. They need to have also done their homework — they need to know how I train people and how they think they would fit in.

When have you made exceptions?

There are one or two cases where I was completely sure that they would get a fellowship, and they didn't. But by then I had gotten so excited about the project we had co-developed that I chose to support them from my own funds.

How does your lab develop a research project?

The ideas often emerge from conversations that start about 18 months before the postdoc comes to work with me. It has almost always been my policy that incoming postdocs build their research projects and are free to take the project with them once they leave, to help them to set up their own lab. That gives the postdocs who are leaving a good opportunity to establish themselves without having to compete with me and the people in my lab.

What careers do your postdocs pursue?

About half the 100 or more postdocs that have gone through my lab hold full-time academic jobs, of which running a research lab is a big component. Many people whom I take on as postdocs want a job in the biotechnology arena. The other 50% are dominated by those who choose to join a company. Those can range from start-up biotech companies to very well-established pharmaceutical or chemical companies. Others lead research groups at institutes or government labs, work as research associates, teach science or do other science-related work. Only two have left science.

Do they get permanent positions right away?

No one who has come through my lab has had to leave science because they could not get a job. Personally, I believe that I have failed a postdoc if I take them into my lab and they cannot get a job that they love when they leave. That usually means that they have to go on to do a second postdoc. There have been very few such individuals — fewer than five, in the 32 years I have had my own lab. So overall I rate my success rate in helping postdocs get their first job at about 90%.

What do you see as the role of a postdoc?

The meaning of postdoctoral training has been lost in today's scientific community. As mentors, we need to really reconsider what we are training postdocs for. And that's just it: it's a training period, not a job.