This project uses natural language processing and computer science techniques to find links between toxic chemical exposures and various diseases that are currently alarmingly on the rise in the United States, and, more generally, in the industrialized world. The approach is multipronged, and it involves integrating information from diverse sources in order to synthesize a story explaining potential links between chemicals and specific diseases. The strategy is to first look for highly significant temporal correlations potentially linking a chemical to a disease or a set of symptoms, and to follow this up with a perusal of the research literature characterizing the disease on the one hand and the toxic effects of the chemical on the other hand. Ultimately, the goal is to uncover a plausible biological mechanism explaining why the chemical would be expected to be causal in the disease, even, in many cases, leading to advances in our understanding of the disease process.
We have made use of multiple online databases readily available for download from the internet, and maintained by the US government, such as hospital discharge data provided by the Centers for Disease Control (CDC) and the FDA's adverse event reporting system (FAERS) as well as their vaccine adverse event reporting system (VAERS). We also use data from the US Department of Agriculture on usage of agrichemicals in food crops. Statistical analyses of word frequencies within these databases can reveal interesting associations between vaccines and/or drugs and/or agrichemicals and specific symptoms and/or diseases.
We have been particularly interested in autism, as this is a debilitating disease affecting children whose prevalence has been rising exponentially over the past two decades. Autism is associated with a complex set of comorbidities, and this provides a rich context in which to explore plausible disease mechanisms. We have been struck by the remarkable correlation between autism rates and the use of the herbicide glyphosate on core food crops, and we are now convinced that this correlation represents a causal relationship.