Tag Archives: openness
Here is a guest post from the awesome Kyle Machulis of OpenYou.org:
Hi, my name is Kyle Machulis, and I like numbers. A lot. One of my hobbies is staring for long periods of time at the numbers that computers and hardware send to each other to figure out what they mean. This is known as reverse engineering. It’s basically my own special, nerdy version of sudoku.
In the past, I’ve mainly worked with video game controllers, such as the Novint Falcon and the Microsoft Kinect. Along the way, I’ve picked up random biometrics hardware here and there. Earlier this year, I realized I’d ended up writing, maintaining, or contributing to drivers for several devices.
- libomron – Data access library for omron pedometers and blood pressure monitors
- liblightstone – Hardware access library for the Journey To Wild Divine Lightstone USB widget
- libfitbit – Hardware access library for the Fitbit pedometer device
- emokit – Raw data access library for the Emotiv EPOC EEG headset
I recently realized it’d be a good idea to create a central place for these kinds of projects, so I started the OpenYou project.
This project is a resource to developers, providing updates and lists of open source libraries for whatever health equipment I can get my hands on, from sports sensors to medical equipment to DIY gadgets. The project aims to get developers to write code that requires them to get out of their chairs to use, and to get non-developers talking to developers about shared needs and ways to refine data transfer and visualization.
So, out of all the hardware available to work on, why concentrate on reverse engineering health and biometrics equipment? The fun part about reversing this hardware is that the numbers you are staring at when figuring things out are yours. While reverse engineering a blood oxygen sensor, you can breathe deeply and watch the data change to map out what things mean. You end up learning about yourself while also learning the hardware.
Outside of the selfish fun of manipulating the data you’re taking apart via physical exertion, there’s the much more important issue of people that really, seriously need the data. Sure, there are those that want to increase their attention or to fine tune some portion of their diet. Then then there are the people that will quite literally die without this data — data that is sometimes locked into a device they can’t get to, a device that will only send it out to a remote company or a certain piece of software, that will send it to their doctor, who will then come back to them with it, when they have time.
That unhelpful practice needs to stop, and we’re here to make that happen. It’s your data, it’s your right to have complete access to it.
OpenYou is happy to take requests to work on hardware that people like access to. The project is also looking for anyone that’s doing sensor development or reverse engineering in an open source manner. There’s a blog to fill, and stories about interesting hardware uses, DIY projects, or other things relevant to open source health development are welcome.
Current projects include:
- Finishing an ANT/ANT+ library for Python (part of the libfitbit project)
- A BodyMedia/BodyBugg library
- DIY projects for the newly opened Zeo firmware
- Gathering information and code off of the Garmin forums to make a central resource for Garmin/Suunto/Polar sensor communication
In April, 2005, the National Institutes of Health and Department of Energy launched a [task force](http://www.genome.gov/10001808) on genetic testing. The task force, which had broad academic and industry representation, tried to outline the issues, and recommend possible guidelines, for the coming age of consumer genetics. It was already obvious that cheap and ubiquitous genetic testing was on its way. The result will be a revolution in medicine comparable to the explosion of pharmacological knowledge in the twentieth century. But of course the complexity of pharmaceutical therapeutics has been mediated by the medical monopoly on prescription drugs. Some drug treatments are simple: have backache, take aspirin. Some are more complicated: have terrible backache, take [Demerol](http://en.wikipedia.org/wiki/Meperidine) ; no, take [Percocet](http://en.wikipedia.org/wiki/Oxycodone); or no, try good old fashioned [morphine](http://en.wikipedia.org/wiki/Morphine).
When the NIH-DOE task force looked at genetic testing, it envisioned an analogous system of mediated access; perhaps not as strictly regulated as the drug system, but still with trained professionals in a central role. “Consumers should discuss testing options with a health care provider competent in genetics prior to having specimens collected for analysis,” wrote the task force in its [final report](http://www.genome.gov/10002401).
“The Task Force discourages advertising or marketing of predictive genetic tests to the public.”
But medical professionals can only ethically recommend tests and treatments that have been proven to be effective, except under explicitly experimental conditions, when patients consent to take a risk and participate in research. Thus [23andMe](https://www.23andme.com), the consumer genetic testing company, does not offer “diagnostic” or “predictive” tests. Instead, the company offers a light dusting of information about the association of traits and diseases with particular genetic variations, along with a tremendously large amount of raw genetic information, in the form of hundreds of thousands of SNPs, or [single nucleotide polymorphisms](http://www.ornl.gov/sci/techresources/Human_Genome/faq/snps.shtml).
Between these two features — a very slim and ambiguous set of potentially significant interpretations; and a big collection of raw SNPs — there is nothing.
The answer is not that 23andMe is being stingy, exactly. Rather, the power of genetic testing cannot be realized until genotype is linked to phenotype. The genetic variation in human beings has to be connected with its effects on our bodies and our behavior. This is the purpose of genetic research. Today, the territory between the genotype and the phenotype remains largely unmapped. Where there is ignorance, there is silence. The fakey quality of the 23andMe GeneJournal is, in fact, a side effect of this clinical silence.
Right now, the important thing about popular genomics is not the diagnostics but the data. 23andMe, along with the other consumer genetics companies, and academic research organizations around the world, are collecting raw genetic information. There are clear indications, from the 23andMe Web site, which I have been exploring carefully, examining the results of my own test, that the company intends to sell the genetic data its customers provide to commercial firms for research purposes. This has attracted some [critical comment](http://www.palidwor.com/blog/?p=63).
It seems to me that there ought to be a different basis altogether for this research. The future of genomic medicine hinges on a model that is neither the “closed” model of highly regulated drug research nor the “closed” model of opaque data-gathering by private firms. Instead, in my utopia, there is a culture of genetic openness. It should be easy to “donate” one’s genomic information to a common pool, and also easy to attach additional information about one’s phenotype. The only thing that needs to be masked from public view is personally identifiable information, the kind of information that would allow particular individuals to be tracked without permission.
This does not necessarily undermine the business model of companies like 23andMe. There is plenty of business to be done in the context of an open culture. But the actual data should be widely and conveniently accessible.
I recently emailed 23andMe to see if there was any easy way to export the hundreds of thousands of SNPs in my “Gene Journal” to a raw data file. This is not because I am a geneticist. Instead, I simply want to know how accessible the 23andMe data is. One of the themes of our exploration here on the Quantified Self is that tracking leads to data leads to experiment leads to self-knowledge. It remains to be seen whether one’s personal genome be an element of this process. The genome is a lot more complicated than many of the behavioral techniques we’ve discussed here. It can only become meaningful through massive collaboration.
At the basis of this collaboration will be a culture of genomic openness.