Two stories recently caught my eye – very different but with a common message that forms the very core of all good science. Volume is everything.
How many times has the mainstream press run with stories of miracle treatments whilst ignoring the thorny issue of sample size. XYZ works because the person who got the treatment got better, the one who didn’t, got worse!
Of course statistically significant experimental design is an essential part of modern research but it’s still a concept many of the public find hard to grasp. Either a drug works or doesn’t – how can there be any other answer? But once you get beyond the level needed for statistical significance you enter another league…
Big data and vaccine manufacture…
The first story, from InformationWeek, was about Merck using “big data analytics” to solve a vaccine manufacturing issue. According to George Llado, Merck vice president of information technology - "we took all of our data on one vaccine, whether from the labs or the process historians or the environmental systems, and just dropped it into a data lake." Presumably a “data lake” is what ultimately happens to the rain that comes from the cloud.
In three months, Merck was able to do 15 billion calculations and more than 5.5 million batch-to-batch comparisons. It discovered a link between what was happening in the fermentation phase of vaccine production and its yields during the final purification step.
Big data analytics have become relatively common in pharma R&D but it is rarely applied to manufacturing problems. Volume data like this can give researchers real confidence in any patterns that they reveal.
Big data and Alzheimer’s research…
"For too long scientists studying Alzheimer's have been working in silos, engaged in a single-minded 'race' to try and beat the disease," according to Prof Julie Williams, principle Investigator of a new £6 million research project at Cardiff University. "That's simply not going to happen unless we pull together.”
The study aims to produce the most comprehensive understanding of the disease's various risks to people by exploring the combined influence of genetics and lifestyle in the development of Alzheimer's in one million people.
Prof Williams’ study will "harmonise the research of scientists studying the genetic risk of Alzheimer's with the work of those studying the lifestyle influences, with the ultimate goal to creating more personalised treatments for the disease and, better yet, treatments that offset it altogether.”
She believes it is going to be large enough to get answers.
Who knows what discoveries could be made by rigorously analysing previously un-pooled data. Of course, applying the principle widely has its difficulties. Data will be inconsistent it the way it was gathered and the characteristics of the participants but get big enough and these differences will become less important.
In theory.
|
|
 |
|
|
Share on LinkedIn |
Share via Email |
