Let us suppose that somehow a not-for-profit organisation has been established to allow anyone that wanted to to sequence and share their DNA profile. I am not interested at this stage in discussing all of the problems associated with that - I would just like you to imagine it.
Let us also suppose that those who are willing to share their DNA are also willing to continuously submit data from their bodies by way of wearable technologies. Again there are complex issues here but please ignore them for now. I just want you to focus on the idea.
We already have a mature market in wearables, particularly in relation to fitness and for a relatively low-cost it is already possible to wear a device that can gather data on
Furthermore there are already low-cost devices that can measure
Input devices are becoming smaller and cheaper all the time and by bridging hardware devices to smartphones via something like Low Energy Bluetooth we suddenly open incredible opportunities.
Already devices like Fitbit and Apple Watch gather data from the body and push it to a cloud platform that analyses and reports back to the user. This gives data scientists the ability to analyse the network of users as a whole and draw meaning from it.
What would happen if we combined DNA sequences with rich and frequent data from the body on a cloud platform? This is the uncomfortable bit. Let us assume that individuals around the globe have opted to submit their data from sensors on their bodies via a smartphone bridge. Imagine that the platform also understands their DNA sequence. Imagine the wealth of data that a network of users with DNA sequences and data from their bodies would present. Imagine the meaning, learning and understanding that could be drawn from it both from research scientists and data scientists.
The platform we are imagining understands the DNA profile of an individual and is gathering large amounts of data from users. Let us suppose that it is possible for users to grant access to their health care professionals. During a consultation a wealth of data and the DNA profile of the patient is available. As medical conditions become apparent users can choose to notify the platform that they have a condition. Perhaps in five years thousands of conditions will be registered on the platform along with user data and DNA profiles.
The resulting platform becomes a self-learning research platform. Machine learning and data science could extract deep meaning from user data and align it to DNA and medical conditions. I would hazard a guess that within a few years a major breakthrough would be uncovered or at the very least deep knowledge around conditions and profiles would be gained.
This is not a deep technical challenge. Wearables, smartphones and big data are already mature technologies. What is a challenge is navigating the moral, political and social issues with this idea. To date massive NGO’s and Governments have driven Global Health programs but we could think differently here.
Blockchain has the potential to protect personal data and decentralise it from any central organisation but in reality this area is nascent and problematic.
Considering the upsides of accelerating medical knowledge and discovery and ultimately humanity’s ability to understand disease and illness my question is this:
Are we brave enough to imagine the idea of a real time global self-learning medical research platform?
It is technically possible.
Have an update or suggestion for this article? You can edit it here and send me a pull request.
Linux and Unix more command tutorial with examples
Tutorial on using more, a UNIX and Linux command for viewing the contents of a file or files one screen at a time. Examples of viewing a file, viewing multiple files, searching using regular expressions and opening the file in a text editor.
Vim: you don't need NERDtree or (maybe) netrw
Vim's netrw file browser is good enough. With a few tweaks there is no need for plugin like NERDtree. For many tasks you may not even need netrw.
Linux and Unix kill command tutorial with examples
Tutorial on using kill, a UNIX and Linux command for terminating a process. Examples of killing a process, sending a SIGTERM, listing signal names and numbers, and handling 'operation not permitted' errors.