Search

In an evolving population, proliferation is dependent on fitness so that a numerically dominant population typically possesses the most well adapted phenotype. In contrast, the evolutionary “losers” typically disappear from the population so that their genetic record is lost. Historically, cancer research has focused on observed genetic mutations in the dominant tumor cell populations which presumably increase fitness. Negative selection, i.e., removal of deleterious mutations from a population, is not observable but can provide critical information regarding genes involved in essential cellular processes. Similar to immunoediting, “evolutionary triage” eliminates mutations in tumor cells that increase susceptibility to the host immune response while mutations that shield them from immune attack increase proliferation and are readily observable (e.g., B2M mutations). These dynamics permit an “inverse problem” analysis linking the fitness consequences of a mutation to its prevalence in a tumor cohort. This is evident in “driver mutations” but, equally important, can identify essential genes in which mutations are seen significantly less than expected by chance. Here we utilized this new approach to investigate evolutionary triage in immune-related genes from TCGA lung adenocarcinoma cohorts. Negative selection differs between the two cohorts and is observed in endoplasmic reticulum aminopeptidase genes, ERAP1 and ERAP2 genes, and DNAM-1/TIGIT ligands. Targeting genes or molecular pathways under positive or negative evolutionary selection may permit new treatment options and increase the efficacy of current immunotherapy.

Background: Basic and clinical scientific research at the University of South Florida (USF) have intersected to support a multi-faceted approach around a common focus on rare iron-related diseases. We proposed a modified version of the National Center for Biotechnology Information’s (NCBI) Hackathon-model to take full advantage of local expertise in building “Iron Hack”, a rare disease-focused hackathon. As the collaborative, problem-solving nature of hackathons tends to attract participants of highly-diverse backgrounds, organizers facilitated a symposium on rare iron-related diseases, specifically porphyrias and Friedreich’s ataxia, pitched at general audiences. Methods: The hackathon was structured to begin each day with presentations by expert clinicians, genetic counselors, researchers focused on molecular and cellular biology, public health/global health, genetics/genomics, computational biology, bioinformatics, biomolecular science, bioengineering, and computer science, as well as guest speakers from the American Porphyria Foundation (APF) and Friedreich’s Ataxia Research Alliance (FARA) to inform participants as to the human impact of these diseases. Results: As a result of this hackathon, we developed resources that are relevant not only to these specific disease-models, but also to other rare diseases and general bioinformatics problems. Within two and a half days, “Iron Hack” participants successfully built collaborative projects to visualize data, build databases, improve rare disease diagnosis, and study rare-disease inheritance. Conclusions: The purpose of this manuscript is to demonstrate the utility of a hackathon model to generate prototypes of generalizable tools for a given disease and train clinicians and data scientists to interact more effectively.