Congratulations to Swati Banerjee Carr!

On Monday at Boston University, I attended the successful Ph.D. thesis defense of Swati Banerjee Carr, who I have been co-advising with Doug Densmore for the past couple of years.

Swati's work is, to my thinking, a classic case of how a lot of science doesn't start with "Eureka!" but with "That's funny..." or, as in Swati's thesis: "Why doesn't any of this work?!?!"  Originally, she was building a simple test system to characterize various different pairs of repressors and promoters in E. coli.  These are some of the most basic and common building blocks in synthetic biology: a repressor acts on a promoter to suppress the gene controlled by that promoter, which means that controlling repressors can switch things on and off in a cell, including other repressors, making it one of the foundational tools for controlling the behavior of cells.  Practically everybody uses them, and in her system she was using some of the most well-known and best understood biological parts around, and yet when she put them together nothing worked.  When she put the parts together in different orders, however, sometimes it worked and sometimes it didn't.  "That's funny..."

With a bit of study, including results of some of my other recent work, the key problem was identified as being the bits of DNA just before the promoters, which changes depending on what else is in the system and how the parts are arranged. And so, Swati's project ended up shifting away from the original plan and focusing instead on solving that problem of context dependence. Now, at the end of her doctoral work, our work together has resulted in a lovely protocol for creating "upstream insulators" that make a promoter behave pretty much the same no matter where you put it.  The data is really beautiful, but I can't share it quite yet: not until Swati has officially deposited the thesis.  What it shows, however, is a remarkably stark contrast:

• Without insulators: a genetic circuit in which every permutation is different, pretty much none of them are "working" by even the most generous definition, and there's not even any real pattern to the chaos
• With insulators: every permutation does almost exactly the same thing, with quite strong and consistent signals for all of them.  

I think this has the potential to be quite a big deal, and don't know why anybody would ever use an uninsulated promoter again once this gets released.

Swati gave a good talk to a packed room, standing room only.  She's got some homework from me and the others on the committee, to improve her actual thesis document some more before it can be considered quite complete, but she's on the home stretch, having done some damned good work, of which I think she can rightly be quite proud.

Reproducibility of Fluorescent Expression from Engineered Biological Constructs in E. coli

I've talked about the interlaboratory studies we've been running through iGEM previously, and while I've been quite excited before, now is the biggest news I have to announce yet: our paper on the results of the studies has just been published in PLOS ONE: Reproducibility of Fluorescent Expression from Engineered Biological Constructs in E. coli.

Fluorescence from iGEM interlab constructs (Credit: Oxford iGEM 2015)

In sum: over the past two years, teams from nearly 100 institutions around the world measured the same simple genetic constructs for expressing green fluorescent protein, and this paper reports their results, crediting all of the more than 600 iGEM authors involved, including lots of undergraduates and even high-school students. In my eyes, the two key results from this study are:

• Ratios between strong fluorescence were remarkably precise.
• Weaker measurements were extremely unreliable, but the problem does not appear to be the biology!  Instead, it appears to be differences in how people use their instruments and handle their data.

These are really good news, because it means that some of the well-known problems in understanding and reproducing biological research might be tackled simply by improving our ability to calibrate our instruments and communicate about our measurements.  That's hard, but it's a lot better than thinking that biology might just be inherently too messy to understand properly.

We've also published all of the raw data submitted by all of the teams, so that people can dig further into the data if they're interested and see what else may be lurking there.

And what will 2016 bring?  That is still in planning, dear reader: you'll just have to wait and see...