Last week, a fellow PhD student and I were remarking how our dissertation topics have been tying into current public and academic discourse. Whether it’s media coverage or peer-reviewed literature, we’re seeing our areas of study get a little deserved time in the limelight. I said that’s why we do what we do: anticipatory scholarship.
What do I mean by anticipatory scholarship?
I blurted out that term in our conversation, and have since been reflecting on it. What is anticipatory scholarship and why is it important? I think of it as researching a topic that at first doesn’t seem to have much traction or attraction, but in time gains salience and relevance based on current events and trends in research. At the start of my PhD, I of course had an argument for why it was important to study people’s access to their raw genetic data and subsequent usage of third-party interpretation tools. And while it was a good argument, the concerns raised admittedly seemed a bit…distant. Direct to consumer (DTC) genetic testing was popular but not that popular, and isn’t it rather niche and quirky to download your raw data and upload it to interpretation websites?
Well, since then DTC genetic testing has only increased in popularity, and more genetics professionals are starting to take note of third-party interpretation (I’ll give two specific examples below). While I didn’t predict these exact events, I did have a sense that lay access to personal genetic data on a large scale was going to become a big deal – or at least a bigger deal then when I started off. Thus the anticipatory nature of my project.
Anticipatory scholarship is important for similar reasons as basic science research. The relevance isn’t always obvious at the outset, but if you don’t invest time and money in possibilities then you will miss out on opportunities to leverage that knowledge when the direct applications become clearer. It’s fair to say all research is in some sense anticipatory: you study something now to gain future insight. But there are some areas that push this way more than others; perhaps it’s the nature of PhD work to be a little less mainstream, a little more anticipatory due to the way it is (or isn’t!) funded.
Obligatory section on Facebook and Cambridge Analytica
Like many, I have been devouring with hunger and morbid fascination the news and commentary about Cambridge Analytica and the abuse of Facebook data. It feels as if we’re all watching our own social media Titanic go down, while also standing on the deck posting status updates on Facebook about how our feet are starting to get wet. For me, the most helpful and informative part of the news coverage is when internet and privacy scholars are invited on to radio spots to explain to us all what has been, is, and perhaps will happen with our hyperconnected world. I silently cheer when they come on, knowing that their knowledge is deep and rooted in years if not decades of research. Now they get to bloom in the sunlight of public attention (well, maybe it isn’t that rosy – but I really do appreciate their expertise!). At a time when it feels we need to brace ourselves against the onslaught of current events, we need solid research and sound science that we can draw up on moments of crisis and upheaval.
DTC testing and raw data in the news
The tie-ins between my research and current events are not on the scale of the Facebook and Cambridge Analystica debacle, but they are important nonetheless. And interesting to compare side by side.
First, an article in a top genomic medicine journal, Genetics in Medicine reported a 40% false positive rate for DTC testing data when sent for confirmation at a clinical sequencing lab. The pie chart in their main figure is indeed scary. I do take some issue with the way this was reported, however, which I outlined in a Twitter thread here:
I have several thoughts about this paper, as my #PublicHealthGenetics dissertation work is about DTC raw data and third-party interpretation services 1/ https://t.co/SCsmfMYdpf
— Sarah Nelson (@blueyedgenes) March 24, 2018
Specifically these two points:
For 26 of the 49 patients, the authors are relying on results transcribed by the very providers for whom there’s concern of “minimal genetic training.” Was there consideration of error in these transcriptions? 6/
— Sarah Nelson (@blueyedgenes) March 24, 2018
Basically, I’ve no doubt that DTC companies generate data with errors, and I know from my study of third party interpretation tools that many rely on annotation databases that wouldn’t hold muster for a clinical geneticist. But I also sense an anti-DTC sentiment that runs through this work, which is understandable given that healthcare is having to pick up slack left by DTC companies when customers want clinical confirmation, guidance, and follow-up care.
In other raw data news (h/t to my advisor), the large Utah-based healthcare system Intermountain Healthcare is announcing a new online registry where patients can upload their raw genetic data from DTC testing. Paired with medical record data, the registry will enable research and potentially be used in making treatment decisions about patients. I find this hard to square with the GIM report above. So DTC raw data is flawed and variant calls are overturned 40% of the time by clinical sequencing. But yet we want to pair it with patient charts and use it for medical decision making. Granted, these are not the same groups saying this, but it points to a fraught area in which I hope to intervene with my anticipatory scholarship. What exactly should patients and consumers be able to do with their raw data, and who should oversee that, if anyone?