The whole idea of actually paying to read mainstream news online is rather alien to me, having grown up immersed in a world of free content readily available via Google News. Indeed, I can’t help but see free news as some kind of inviolable human right.
Thus when the Times recently set up a paywall blocking free access to all of its online content (including its blogs), I was faced with a serious dilemma: there are only a few mainstream science journalists in the world who write sensibly about genetics, and my favourite example (Mark Henderson) was now locked away behind a web-form requiring my bank details.
I resisted as long as I could, but two weeks ago I caved: for the first time in my life I’m actually paying for online news, and it’s basically entirely because of Mark’s articles.
Here’s an example: Mark had a two-page spread in the Saturday edition of the Times (article here
, background piece and fancy infographic here
, and related blog post here
; all require a subscription) looking in detail at why his predicted disease risk varies between the three different personal genomics companies he’s had his genome scanned by (23andMe
, Pathway Genomics
). It turns out in many cases this variation is driven not by inaccurate data, or by the use of different markers, but something far more mundane: differences in the companies’ estimates of the background (general population) risk of the disease.
Because the companies use this number to calculate a customer’s individual absolute risk of a disease, that can have a huge impact on risk predictions: it results in Mark’s predicted risk of exfoliation glaucoma being almost 18 times higher in his deCODEme results than his results from 23andMe.
This problem isn’t a new one: it was raised back in July 2008 at a meeting between representatives from 23andMe, deCODEme and Navigenics to discuss industry-wide standards (see report here
), but has never been resolved. There are valid reasons for companies to give differing predictions for the same individual, but this isn’t one of them; the fact that problems like this have persisted for over two years is a serious black mark against the industry. Stronger, earlier moves towards self-regulation to address these types of issues may have helped to head off the regulatory crackdown currently looming over the industry
As an added bonus: if you do decide to cross the paywall to read Mark’s coverage, you’ll also find a brief op-ed
written by yours truly riffing on Dan Vorhaus’ argument
that such problems are best resolved via transparency rather than stringent regulatory obstacles. (And yes, paying to see my own article was a somewhat disconcerting experience.)
Ayway, kudos to Mark for taking the time to dig deeply enough into his own genetic data to track down these problems and their underlying basis, and for reporting on them without resorting to hyperbole about dangerous, misleading results.
Added in edit: paying for a Times subscription is made easier by the fact that there’s currently a special price of £1 for the first month on offer. Once the price rises to the proposed final level of £2/week it will become substantially harder to justify, even for someone who enjoys Mark’s articles as much as I do. Times accountants, take note: you’ll get a lot more money flowing in if you allow people to pay to access specific content (e.g. specific articles via micro-payment, or reduced subscriptions for purely the science content) than if you force people to pay for a vast amount of stuff they’re not interested in reading.