Revealing the Factors of Disease in Big Data

Just a few weeks ago, a jury in San Francisco awarded plaintiff Dewayne Johnson $289 million in his claim against Monsanto (now Bayer) that the herbicide, Roundup® caused his cancer. Glyphosate, the “probable carcinogen” as designated by the World Health Organization has been commercially sold for almost 45 years and has become the most widely used herbicide ever. It is probably accurate to state that our ability to “feed the world” during the latter half of the exponential human population growth of the 20th century was greatly facilitated by the application of glyphosate across the planet.

Although decades of research have been conducted on the safety profile of the chemical, a persistent disagreement across the scientific community continues whether or not glyphosate can cause cancer in humans. Much is known about glyphosate’s biological activity and mechanism of action (1). What has changed? Why now and what do we, as concerned consumers need to know about the effects of exposure to this chemical?

While it takes years to conduct controlled research studies to identify causes of complex diseases like cancer, we have access like never before to a massive body of scientific knowledge, and we benefit from global connections that allow us to spot patterns in aggregated data. Single case studies, inherently anecdotal because they are rare, become handfuls of grouped observations, and multiple research reports can now be analyzed in “meta-studies” to enrich our understanding of causative relationships. Thus, while individual studies may provide conflicting results, aggregated and disseminated data may reveal statistically significant associations like chemical exposure and cancer that can point to areas in need of a more in-depth investigation (2). Even where such data challenges the existing scientific understanding of a disease and its potential causes, the possibility of finding previously undiscovered associations in “big data” is incredibly important to appreciate. Indeed, this verdict is actually the second this year where a jury has decided that enough evidence exists for a link between a chemical and the cause of a specific cancer despite the question of causality being inconclusive in the scientific literature (3)(4).

For the majority of consumers who do not have the experience to evaluate research findings like those conducted on glyphosate, finding a trusted source of interpretation and the tools to provide confidence in the analyses will be crucial for informed decisions. Perhaps it will be the ability to represent knowledge across scientific domains like chemistry, biology, and human disease. Alternatively, maybe it will be the application of A.I. To detect hidden associations in massive data

sets. Perhaps it will be “all of the above.” At the end of the day, providing analytical certainty in evaluating all of the information available to us will be critical for informed choices in this information-based age.

References
  1. https://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=3496#section=Top
  2. https://www.nature.com/news/widely-used-herbicide-linked-to-cancer-1.17181
  3. https://www.nytimes.com/2018/07/12/business/johnson-johnson-talcum-powder.html
  4. https://www.cancer.gov/types/ovarian/hp/ovarian-prevention-pdq#link/_220_toc

Posted on

August 20, 2018

in

AI

category

related

100+

Years of Experience

99%

Positive Feedback

many

Sleepless Hours

1000+

Cups of Coffee