One of the most important factors in assessing disease, especially cancer, is accurately distinguishing among the possible variants. This can be a difficult task, however, especially when the best tool for the job is a microscope. Over the last several years diagnostic tools and criteria have been improved by combining large scale data processing with various molecular characteristics.
The work in the linked article accomplishes this for 100 types of neurological tumour. Furthermore, they used DNA methylation, a type of epigenetic modification, to do it.
The significance of DNA methylation patterns is two-fold. First, scientists have learned that these patterns can be very good predictors of both traits and disease. These discoveries have led to the explosion of epigenetics research in the last decade. Second, molecular toxicologists have worked on elucidating the other side of the equation, linking chemical exposures to epigenetic modifications. Some of these results are stored in publicly accessible databases.
Should a set of DNA methylation changes be linked to both a chemical exposure and a specific tumour sub-type, the patient suffering from the tumour suddenly has the evidence needed to become the plaintiff in a lawsuit seeking to recover damage from those responsible for the chemical exposure. If enough patients can demonstrate this link the conditions quickly become ripe for mass litigation. Many of Praedicat’s mass litigation scenarios explore the consequences of exactly these sorts of scientific change, giving our clients the ability to manage the risk.
Our results provide a blueprint for the generation of machine-learning-based tumour classifiers across other cancer entities, with the potential to fundamentally transform tumour pathology.