DNA – Data Not Amendable

I’ve always been intrigued by sites like 23andMe and AncestryDNA which give you insights based on a DNA sample. They claim to be able to tell you about your family heritage, identify other users with a close DNA match, and potentially indicate whether you’re susceptible to certain inherited diseases.

But I’d never got around to signing up for one. They’re not free; usually you have to pay for a kit. Also, I can’t imagine that I would get much useful data from them either. Ethnically, I’m White English, and as far as I know, so are most of my family tree going back several generations. Indeed, we haven’t really moved much out of the north of England. So such sites would be unlikely to tell me anything that I didn’t already know.

As it happens, recent events have made me even less inclined to sign up. A smaller UK firm, Atlas Biomed, seems to have gone bust, with no indication of what will happen to the data that users have provided. And the aforementioned 23andMe is in trouble; it’s cutting 200 jobs, representing 40% of its workforce, and seems to be losing multiple millions of dollars. Oh, and it got hacked last year.

Your DNA is not something you can change. It’s not like a bank account; should someone gain unauthorised access, then it’s possible for your bank to give you a new account number and cards. You can change email addresses, and, under witness protection schemes, even gain a whole new identity if needed. But you can’t make wholesale changes to your DNA (gene therapy notwithstanding). And so if your DNA leaks out, or is sold on, there’s not much you can do about it.

If you’re reading this and are thinking about using a DNA testing service, maybe undertake some due diligence first. Ask questions like: where is the company based, and where is it storing your data? How will they keep it secure? Will they securely delete all your data if you withdraw consent?

Big data and data analytics Seminar

A photo of the interior of the Glazier's Hall

On Thursday, I was lucky enough to be invited to attend Big data and data analytics: commercial opportunities, privacy and effectiveness, one of several seminars offered by the Westminster eForum. It took place at the Glazier’s Hall, on London’s south bank next to London Bridge.

The four hour session, split into two halves, was chaired by two members of the House of Lords, Lord Inglewood and Lord Witty, and the speakers represented various users of big data in the UK. These included the ABPI, whose members carry out research and development into new medications, Dunnhumby, who worked with Tesco to launch the original Clubcard in the 1990s, academics and industry partners.

The talks given by the speakers were interesting, and focussed more on policy and high-level overviews, rather than technical details. For example, whilst Hadoop was passively mentioned on some slides, there wasn’t much about deployment and how it works. But there was some discussion about database design, as companies move away from traditional relational databases to big data capturing solutions.

Privacy implications came up several times as well, an irony not lost on one of the speakers who noted that the event coincided with Data Privacy Day. In particular, there was a focus on how to design systems with privacy in mind, but also that the UK’s and Europe’s more restrictive privacy laws may be part of the reason why the world’s biggest data users – GAFA (Google, Apple, Facebook & Amazon) – are all based in the US.

I came into the seminar essentially wearing two hats. My main reason for attending was as a blogger (or ‘freelance writer’ according to the attendance list), but data analysis is also one of my roles at work. However, we’re not yet at the stage where we’re using ‘big data’ – most of our data is all within standard relational databases and I can’t see that changing any time soon.

As always, such events also offer a chance to network and it was good to speak with some of the other attendees. As you’d expect from a more high-level seminar, this was an event for people with suits and ties, and not t-shirts and hoodies. Many were from government departments, regulators and other public sector bodies, as well as large organisations such as the BBC and Arqiva.

I came away with plenty of notes, and some action points to perhaps bring up at work. Channel 4’s Viewer Promise video was mentioned as great example of best practice for explaining their privacy policy – far better than pages and pages of legalese. Maybe universities could do something similar to explain the student contract at enrolment.

This day was made possible by Dell, but all thoughts are my own.