So, there is life after a high-throughput screen after all.
As the dust settles after publication of my big screen in the Journal of Cell Biology, I’ve been thinking back on the last four years of my lab life and wondering where it all went. I knew when we embarked on the project – a visual survey of cell shape phenotypes across the entire fly genome, and in a targeted subset of 500+ human genes – that it was ambitious. But I’d had a real mental block about envisioning its aftermath.
I remember having particular difficulty in my Wellcome Trust fellowship application, in the section where I had to describe long-term aims and milestones for when the screen was “finished”. It’s really hard to plan experiments when you don’t know what sort of genes will shake out of the tree: you are reduced to banal generalities in the subjunctive mood – if I were to get X as a hit, I’d do Y to follow it up. Fortunately the Wellcome panel of experts were sufficiently convinced that the screen would yield good fruit, no matter how hypothetically, and by and large I am happy with the outcome as well.
But there is a strange feeling when you come to the end of such a long project. The momentum that carried you up to that point suddenly dissipates – you are both lighter, but also newly unstructured. In my case, the sense of being cut loose is exacerbated by the imminent end of my contract at the close of January. I think I have found the means to stay in research afterwards, though until the funding comes through and the contract is signed, I don’t want to go public about it. I’m very optimistic about the appointment, but by its nature, I will probably be leaving behind the work I am doing now.
So what do you do with a few months? What can you do? I’ve turned to a second screen – the product of an EMBO short-term fellowship sabbatical I carried out for a few months at the EMBL a few years back. The dataset comprises two terabytes of timelapse video footage of each and every one of those 500+ human genes manifesting their actin phenotypes in real time. After having failed to find any collaborators who had the time and money to help me analyze this massive dataset computationally, and knowing that I couldn’t physically watch all the movies myself, I came up with a cunning plan: to use my fixed J Cell Biol screen to identify one phenotypic cluster of interest and then visually inspect the movies for all the genes in that set to see if anything looked intriguing.
Dear reader, it worked a treat. I have found one gene that gives an amazing and unique phenotype in dynamic timelapse that wasn’t evident from the fixed screen. Like the best hits, there is a little known about it, but not very much. I’ve also got in touch with the researcher who’s published the sole paper on this gene, and we’ve embarked on a really friendly and stimulating overseas collaboration – I don’t want to reveal any details just yet, but we have high hopes of getting a paper together. It’s been so long since I’ve worked on just one gene that I’d forgotten how good it feels to abandon the general for the highly specific. When you work on one gene, instead of hundreds, you can make and test mutants – you can lavish care and attention onto your approach; you can read the literature in detail and allow yourself the indulgence of entertaining a few crazy ideas.
Will I ever do a screen again? Quite probably. But for now, I’m basking in the glow of reverting to my previous one-gene-sort-of-girl phenotype at long last.
“it worked a treat. I have found one gene that gives an amazing and unique phenotype in dynamic timelapse that wasn’t evident from the fixed screen”
woho!! Congrats on happy findings and the collaboration! It’s so good to read about those happy times when things pan out well.
Good luck with the next months and look forward to the next chapter/adventure of research!
This sounds wonderful, really good to hear. Enjoy expressing your one-gene-sort-of-girl phenotype 😉
Well done Jenny – isn’t it nice when something small and neat emerges like a pearl from a great big unwieldy oyster? Makes it all worthwhile.
Thanks, all. Of course with science, what looks like a pearl might very well turn out to be a bit of shiny paste. But fingers crossed…
See? Sometimes things just work themselves out. 😀
That is excellent. And getting back to plowing through all the data… I wonder if there is any way of crowd and/or cloud-sourcing the analysis? Even with non-specialists looking at it, if it could be divvied up into little, manageable pieces perhaps it could be done. You might miss things, or have a lot of spurious false positives, from not having one dedicated specialist analyze it all, but it might work.
Just a thought, I realize this sounds like a project all on its own. In the meantime, enjoy your return to the good old days of working on “your own” gene. 🙂
That’s the great thing about screens – you get to work on the big picture and then the nitty gritty details, all in one project! There are a few similarly structured projects on the go in my department, and I hope they bear as much fruit as yours!
Richard, we actually had talked about whether a citizen science approach might work for my data. I’d be up for it in principle, but how to go about the infrastructure?
Also, the users would have to really immerse themselves in the wild-type – it is amazingly heterogeneous.
Yeah, I have no idea how you’d implement it… just pie in the sky thinking. It would be a thing of beauty if possible though. Imagine writing the paper, and describing in the methods section how the data were analyzed. 🙂
Finding someone with both time and money would be hard – but it’s a shame you couldn’t find someone all the same. I guess you’d have had to find someone willing to apply locally for funds via a partnership grant or whatever – ?
I have a related problem with regards research projects (as opposed to working for clients) in a sense – I have the time and skills available (20 years experience in computational biology, etc), but not the funds to do enough with the research projects (my own research interests, that is). It’s very frustrating.
Sometimes you feel like it all comes back to money!
Stephen has just given me the name of a citizen science person here in the UK…I might actually have a chat. It sounds like a lot of fun.
Excuse my whinge earlier – I should know better than write late at night when I’m dead beat after a long day working. It is a shame you couldn’t find a partner, though. It seems to me that there are an awful lot of datasets that are—like yours—waiting for a computational biologist to explore, or explore further. Funding permitting, of course…! :-/
Good luck with the citizen science route!
I have some idea of what you mean about the single-gene thing. A computational counterpart (IMHO) would be exploring mechanisms, e.g. molecular modelling, etc.
It would be the bee’s knees if we could come up with something between SETI@home or FoldIt for analyzing the imaging data associated with high-throughput screening.
You know how I feel about bringing science to the masses.
Personally, I’d like to say thank you publicly because you’ve made science more fun for me than it has been in a long, long while.
Maybe at some point you and I could share more about how we ended up sharing this work?
-Dr X
It will make a great story! Who gets to play us in the film version?