Abstract

Multi-step disease trajectories are key to the understanding of human disease progression patterns and their underlying molecular level etiologies. The number of human protein coding genes is small, and many genes are presumably impacting more than one disease, a fact that complicates the process of identifying actionable variation for use in precision medicine efforts. We present approaches to the identification of frequent disease trajectories from population-wide healthcare data comprising millions of patients and corresponding strategies for linking disease co-occurrences to genomic individuality.

The talk will present an analysis of all significant disease associations occurring prior to cancer diagnoses. Across 17 cancer types, a total of 648 significant diagnoses correlated directly with a cancer, while 168 diagnosis trajectories of time-ordered steps were identified for seven cancer types.

By exploring the pre-cancer landscape using this large data set, we identified disease associations that can be used to derive mechanistic hypotheses for future cancer research. We find that common diseases shared across cancer types converge towards the common theme of chronic inflammation. The trajectory concept can possibly also be used to systematically redefine phenotypes as longitudinal patterns. This can lead to a new way of assessing the validity of diagnoses (mis- and over-diagnosis), or alternatively used to suggest missing diagnoses (under-diagnosis), from their temporal context. Such a diagnosis “clean-up” step is also relevant in conventional case-control studies where false negative and false positive individuals bring down the statistical power when identifying disease related genomic variation.