The view of cancer as an evolutionary problem is surprisingly new: A great majority of cancer studies are conducted either on humans or on artificially induced cancers in murine or other small-bodied mammal models. Two consequences arise from the increasing recognition that cancer is a fundamental property of all parts of the tree of life where the transition to multicellularity has occurred. First, there is the need to understand cancer much more broadly in the wild, to yield insights that could ultimately lead to better prevention and/or treatment in humans. Second, in basic eco-evolutionary theory, cancer risk is traditionally ignored as a factor in life-history evolution, most likely because cancer in the wild typically kills indirectly, by increasing death rates from other causes (predation, parasitism). The evidence that cancer can be a selective force comes from Peto’s paradox: extrapolating from cell-level phenomena, it predicts that large animals should not exist (cancer should kill them before maturity is reached). Yet there appears to be no positive correlation between body size and cancer risk. Data are, however, haphazardly collected. Our aim is to proceed towards an integrative understanding of cancer risk as a phenomenon tied to cell replication and to selection acting on life history patterns in metazoans. We do this in two ways – through studying wild rodents, and by conducting a broad comparison of cancer risk across animals protected from predation.
PhD-Student: Yağmur Erten
Most adaptive traits are quantitative traits affected by many genetic loci of small effect. They are thus often referred to as polygenic traits. Examples are human height, or human disorders, such as Alzheimer's disease or diabetes. Their polygenic basis has important consequences when trying to detect signatures of selection at the underlying genes. Selection on polygenic traits likely acts on standing genetic variation and causes small allele frequency changes at many loci, too small to be detected. Unfortunately, most methods aimed at finding loci under selection still work within the paradigm of a single mutation of large effect sweeping to fixation. Recent genomic studies have had very little success in finding genome-wide patterns of selective sweeps, especially in humans. Furthermore, it is widely accepted that most genes controlling quantitative traits do not act independently but interact through biological networks. Novel approaches try and map the causal interacting variants on gene networks with trait or environmental variation. This shift from single locus to polygenic approaches is still in its infancy but is strongly advocated by the community. With this project, we propose a dual computational and genomic approach. We will simulate DNA sequence data for polygenic traits with additive and epistatic loci. The data will serve to (i) understand how epistatic genotype-phenotype maps affect the evolution of quantitative traits, (ii) test statistical methods for detecting polygenic signals of selection among populations. The genomic part will consist in searching for signals of polygenic adaptation of Sphagnum magellanicum along environmental gradients using NGS data.
Postdoc: Champak Beeravolu Reddy