When I was a kid we had a game called (I think) Frantic Frogs. You wound up a bunch of little mechanical frogs & placed it's rear under a disk in the middle of the board. Once everyone's frog was on the starting line, you lifted the disk and the frogs started hopping towards the outer rim of the board. You had to steer your frogs by gentle taps with a provided plastic stick -- first one to corral all their frogs wins.
Many biomolecular interactions are like those frogs, constantly bouncing up-and-down. This represents a challenge for trying to capture these interactions -- any condition that reasonably washes away what didn't bind is also likely to wash away these transient interactors.
An article a week back in Science provides a clever microfluidic fix to this problem -- by supplying that central disk. In the scheme, an array of DNA spots is bound to a transcription factor. A pressure activated membrane can then drop on each spot, squeezing out the solution but pinning down the solid phase bound molecules.
The article throws in some other tricks: the microfluidic device is actually layered atop a spotted DNA array & the transcription factors were synthesized using in vitro transcription-translation. In short, this
One reason I find microfluidics so fascinating is that most of our everyday understanding of how things work get thrown by the wayside; because the scales are so small, forces such as surface tension take over from things such as gravity. This one stretches the mind again -- one isn't used to thinking about using mechanical forces to keep things from going into solution.