I’m just back home from the Drosophila Research Conference in San Diego, thrilled that FlySorter was received so well at the meeting. Even though we’ve been around for a few years, this was our first appearance at the conference as a vendor, and our first concerted effort to spread the word about what we’ve built. Of course there are fly labs that wouldn’t find our products useful, but even those researchers agreed that there was a need for high-throughput phenotyping tools.
We also shared the news with some of you who stopped by our booth that we had submitted a proposal to NIH for a small business innovation research (SBIR) grant earlier this year. A couple weeks before the conference, we got our scores back, and, well, we’re quite pleased. It’s far from a done deal, but it does look promising that we’ll get some funding to accelerate development of the next step in the sorting process: classification.
Our image capture hardware automatically takes high-resolution dorsal and ventral views of every fly in a vial (like the ones above), and we need to train an algorithm to differentiate, say, a male fly from a female fly. We’re quite confident that the photos are more than good enough to get us to our stated goal of 99.5% accuracy, but there’s work to be done to get there just the same. After that, we’ll develop a hardware module that can direct flies into one of several vials, and then we’ll have an end-to-end sorting system. A vial of flies goes in, and two (or more) vials come out, with the flies separated by sex, eye color, or other phenotype.
If you want to keep up with our progress, I’d encourage you to sign up for our mailing list. (We won’t ever sell or give away your information, and of course you can unsubscribe at any time.)