I can kind of understand that logic with Huntington's but I'd be surprised if it applied to recessive traits as well.
10% of people who used service discovered they have BrandonWirtz's Disease.
Requirements: Male. Primarily Caucasion as primarily limited to Arian Descent. Symptoms onset at 55. Not symptomatic for those with BMI under 22.
With this information I can take the 100% and cut the list to 50% that are male. 50% that are caucasion. (25%) Would have already know if they were over 55 so we cut another 25% Strip last names for ones that would not be of the right heritage. Suddenly the eligible candidates are only 12% of the total. I know that 80% of those are the people on that list.
That's how mining works.
You can't share data anonymously becuase each facet of data allows me more information to piece things together.
Sharing has to be opt in. But if it is opt in then it is no longer statiscitally relevant. Because just as an example. People are going to be less apt to share data if they have a disease associated with a stigma. Which will skew the results.
[Brandon Wirtz's Disease is an imaginary disease used for demonstration only. I don't have any genes shared with other humans as I am from another planet.]
Often a single piece of data will cause the intersection of two traits to be 80% accurate. You can't mathmatically predict these things because you don't know how many factors I have.
There are zip codes where only 1 in 100 people are of a given race. There are diseases that only 1 in 100k people have done a search for and only 1 in 500k people have. These kinds of things you can't formulaize.
| With this information I can take the 100% and cut the list
| to 50% that are male. 50% that are caucasion. (25%) Would
| have already know if they were over 55 so we cut another
| 25% Strip last names for ones that would not be of the right
| heritage. Suddenly the eligible candidates are only 12% of
| the total.
| I know that 80% of those are the people on that
| list.
What list? Are you somehow taking a list of names from outside and correlating them based just on 'X%'?If you have seen those ads on FB that are for something you just searched on Amazon, or that want to sell you hosting for a place you already have hosting with you know what I mean.
If you have enough data about the factors that are common of people with a disease, and then compare that with percentage of people known to have that disease from a set.
If 10% of Company's American clients have Disease X, and Disease X requires you live in a cold climate to be diagnosed ( Raynaud's for example ) then you can exclude the people who live in a warm climate.
Now if you want to target people with Raynauds and you know 80% of the clients live in California,Texas, And Nevada. You know that 50% of the remaining clients have Raynaud's.
This is of course a contrived example, and uses much less data than would be used by a Data Miner.
For the people saying "no there are formula's we can prevent this from happening" yeah you know who made those formula's? The People who make money selling data. And thos formulas are to keep you from thinking it is dangerous.
Counsyl definitely falls under the scope of HIPAA; we do tests at the request of physicians, so we're like any other lab as far as privacy is concerned. Your employer, parents, or prospective mate would only know about the results if you decide to tell them.
It's also important to note exactly what Counsyl is testing: we're look at severe single-gene disorders, things that in most cases cause severe disability or early death. Importantly, if you only have one defective copy of a gene for which we test, you won't show symptoms of the disease. This is not the kind of thing that should affect your employer or insurer. And, if you're a carrier, it's probably something you should discuss with your prospective mate: if he or she is also a carrier for the same disease, there's a 1/4 chance of your child having that severe condition.
*edit to add the last paragraph