My last post discussed some options we considered when deciding the platform for developing a mobile product. I laid out the pros and cons of each option, but left off on a bit of a cliffhanger. The big reveal: we chose to use Appcelerator Titanium. We wanted to be able to have both iOS and Android versions of the product, yet we knew that we'd have to make to do with a small team that could not hope to maintain two separate native codebases. At the same time we really wanted an optimized, native mobile experience that we did not believe could yet be achieved with a mobile HTML5 application. Given these considerations, Titanium was really the only tool that fit the bill.
We've been getting up to speed and working with Titanium for a couple of months now. I wanted to share a few observations about our experience with the various components of the Titanium platform.
The Bio-IT conference itself has grown (reportedly up to 2000 registrants, double from 2 years ago) and in walking the aisles this year, I was pleased to see both familiar faces (colleagues, clients and competitors alike) and new entrants. The world Bio-IT represents and has garnered interest from has also grown and is no longer niche. The initial impetus for us creating 5AM - bringing professional software engineering to the life science domain - has evolved into a more strategic focus coinciding with the broader convergence of life science and health care that provides both concern and hope. The concern is that the convergence doesn't create enough value fast enough to overcome the tension across entrenched cultural, political and business interests. The recent stimulus packages sponsoring HIT adoption and biomedical research and development could go for naught. The hope is that the overwhelming need to reduce the cost of health care, drug development and the precision of care along with the research (and software) required to make it possible will become a focus and a benefit for the industry and the country.
I was fortunate to be offered the chance to speak to a crowd of mostly biologists and researchers (only two vendors in the audience of 60 or so folks!). The purpose was to discuss new collaborative models for science and business that caBIG® fosters and I took the opportunity to discuss open source's place as well. If you're interested, you can download the presentation here, but in essence, during the 15 minutes allotted, we identified the major open source projects employed from a domain independent perspective (e.g., Linux, JBoss, MySQL) to those that are domain specific, such as caBIG® and i2b2. Walking through the principles about open development and the resulting transparency created, the idea that data is more valuable when available to other systems and the generous business friendly license model, underscores why it's an enabler of innovation. We have benefited from both serving the direction of the NCI and being able to pull from the tool belt provided to support customers who were trying to push the boundaries of the status quo with limited means. Being able to reliably reuse components, create reciprocal value, invent where there's need, and focus on delivery of value (versus lines of code) further instills hope. We're not alone in using technology to help converge science and medicine to benefit people. Some days, that 2AM does feel like 5AM. Bring on the dawn.
In case you have been living under a rock since January, SNPTips is 5AM’s popular new Firefox browser extension that allows customers of personal genomics services (like 23andMe) to look up their genotype information while browsing web pages (like journal articles or blog posts) that mention SNPs. It’s a simple idea, but really handy. Since the launch, SNPTips has garnered users in over 50 countries worldwide, and has been reviewed favorably by Daniel MacArthur from Genetic Future at WIRED Science Blogs and the good folks at OpenHelix.
Not ones to rest on our laurels, we at 5AM have been busy improving SNPTips, and we’re just about ready to release a beta of SNPTips 1.1. Here’s what you can expect in the new version:
- Support for deCODEme data files: We’ve added support for deCODEme raw data, so it’s not just for 23andMe anymore! We hope to continue adding support for other vendors in future releases.
- Reference Mode: So, you say you don’t have any genotype data yet? Not even willing to take up 23andMe on their new pricing, at $99 down, $9/month? No problem - you can still use SNPTips to look up information about SNPs on web pages (it just won’t show you your genotype at the SNP). You can still access all of the handy reference links, however, like dbSNP, SNPedia, and Google Scholar.
- We revamped our markup model to make it even more secure: Now, we don’t even add genotype information into the Domain Object Model (DOM) until the popup is shown, and we remove it immediately, leaving the page in memory with no genotype information in the markup.
- We fixed a few other cosmetic bugs and minor issues.
We’re just putting the finishing touches on the 1.1 release now, and we hope to have it out by the end of April. Planning on the next release has already begun, so if you have a favorite feature request, don’t be shy! Contact the SNPTips team at snptips@5amsolutions.com and make your voice heard. High on our list are a Google Chrome version, and support for showing results from multiple data files at once (e.g., family members). These are our two most-requested features so far. As always, let us know what you think, and stay tuned for updates on the 1.1 release! For more updates on SNPTips, follow @SNPTips on twitter.
A final note, in honor of DNA Day - Will FitzHugh, our Chief Science Officer, and I have been accepted into the Personal Genome Project (as part of the PGP-1K) and we have chosen to upload our 23andMe data for public access while we await sequencing. If you’d like to try out SNPTips before you have your own genotype data, feel free to try it out with ours (find the Genetic Data section near the bottom, and click the Download link):
If you find anything interesting in our data, please let us know. Happy DNA Day!
The U.S. Food and Drug Administration recently held a public meeting of their direct-to-consumer genetics advisory panel (agenda here). Luckily it was held in Gaithersburg, Maryland, ahout 15 minutes from where I live. For more complete coverage of the events see the twitter hashtag #FDADTC and Dan Vorhaus' blog posts.
