Enter the Lab

So, there I was, pipette in hand, doing actual labwork for the first time in a year. How had it come to this? When I started out, I was convinced I was not going to be one of those PIs who is never in the lab: a common sentiment if you speak to late stage postdocs/ early stage PIs. My initial determination to stay lab active was in part caused by the disconnect between the training you receive as a post-doc and the reality of being a PI. Labwork so dominates the life of the postdoc that it is hard to imagine a job without it, skewing your sense of what a PI does and should be doing.

It can be tricky to come to terms with but as you progress in academia your job is no longer at the bench. There is a blunt reinterpretation of Adam Smith’s division of labour: “why have a dog and bark yourself”, which means you can’t do everything. As the leader of the group, your main responsibility is to support your team through ideas, funding and papers. Whilst it may be possible to have ideas and troubleshoot from the lab bench, there are some things that cannot be done, particularly writing. I find it very hard to alternate between bench and desk – and when I do try, fail to do both.

As time progresses, I have gotten to the point where my presence in the lab raises eyebrows and prompts various sarcastic comments along the lines of “labwork eh, try not to break anything”. Which I have to answer with as good grace as possible. However, there was one great occasion, when I had been teased all day by a postdoc about PIs not knowing anything, after several hours of this “banter” the same postdoc made a truly basic mistake in their experiments whilst my experiment went as planned, much to my amusement.

The move out of the lab comes with downsides. The lab is the heart of the group: it’s where all the gossip happens. Long experiments are great opportunities to get to know the team. Ensuring you have shared time together through other activities e.g. tea breaks, lunches and socials can help. As you step away from the bench, there is an inevitable skill fade, both in the techniques you do know and newer techniques that you do not. Losing lab skills is problematic on a number of levels. Psychologically, our success as postdocs was so closely linked to our success in the lab, losing the lab skill set before mastering the PI skill set is tricky. It can also affect our ability to lead in terms of legitimacy as group leaders which is in part based on our technical expertise, when this fades it can increase the ever-present imposter syndrome, but also not knowing how a technique works limits the ability to give feedback when it doesn’t. The best solution to this is to hire brilliant people.

There are still times when you can justify your presence in the lab, particularly in training. When you are getting established and it is just you and one or two other members of staff, the majority of lab know-how resides in your head. At this early career stage, there is a tension between good training and good results. You need to build a solid platform, so that your staff can function independently in the future: but every second you are not generating your first paper feels like wasted time, especially with the probation/ fellowship/ tenure clock ticking down. If you survive this first stage, the training can be self-perpetuating with existing staff training the newbies. This is very satisfactory but comes with the caveat that you need to think about quality assurance. The techniques I taught to my first PhD students 10 years ago have morphed over time. Normally this is for the best – things change after all, but it is worth checking occasionally. The training role never entirely disappears, because people selfishly leave from time to time and unless you have a succession plan in place take all that knowledge with them.

It is really hard to strike the balance. There are times when you are best out of the lab and other times when your presence in the lab might be the key difference between success and failure, or at least helping limit the levels of stress that your team have during bigger experiments. Erring on the side of absence builds independence in your team much more quickly – though this can be tough on them. One approach is to treat yourself to some labwork every now and then. Especially if you pick something that you can still do, that generates quick, easy results for use as preliminary grant data. These short bursts in the lab are refreshing. Unlike a PI’s day which involves some meetings, thinking a bit and maybe some writing, labwork has a set timetable with a clear endpoint to the day. Labwork can be simple and clean and it is reassuring to still be good at something, since a lot of being a PI involves external forces telling you are rubbish. It can also remind you why you got into science in the first place, especially if you generate some novel data. But lab time is best treated as a luxury rather than a key part of the job.

There are also upsides to leaving the lab behind. We tend to paint our lab time in a rosy shade as the best time of our life. But if you go back for longer than a day or so, you remember that it can be frustrating, especially when things don’t work. But more importantly, being the head of a lab enables you to do more of the research you want to do. As an early career researcher, you are working on someone else’s project and there is just one of you. When you get your own lab, you can put as many people as you can get funded onto the tasks you want to do. You also spread your losses, if it is just you, when an experiment fails it is devastating, when there are 5 people working for you, any one failure is offset by other successes.

The fact is that your role changes. As your group grows there is less need for you to be in the lab (and this is ok). You have to come to terms with the fact that most of the time you are best serving yourself and your team outside the lab. So if you are not leading a group yet, enjoy the labwork while you can because, as odd as it may sound right now, one day you will miss it.

This article first appeared on Digital Science