The evolution of multicellularity required the suppression of cancer. With every cell having a likelihood of undergoing carcinogenesis, it follows that larger, longer-lived animals must have an increased risk of cancer, compared to smaller, shorter-lived animals. However, observation of natural animal populations contradicts this claim. This discrepancy between observation and theory is known as Peto’s paradox and, despite a renewal of interest in recent years, a quantitative solution has remained elusive.
I developed a novel computational model to test the veracity of Peto’s paradox, revealing a quantitative solution that can explain rate of cancer incidence at the organ level. The model, based upon coalescent theory and incorporating the scaling effects of metabolic rate on stem cell number, cellular generation time and, by extension, mutation rate, was used to predict numbers of tumour-suppressing gene copies required to suppress colorectal and lung cancer in natural populations, and further for a whole-organism. Predicted gene values were compared with known copies from available sequenced genomes - of which three; the orca (Orcinus orca), minke whale (Balaenoptera acutorostrata) and bowhead whale (Balaena mysticetus), were analysed for the first time. Utilising predicted tumour-suppressor copies, it was possible to predict cancer incidence in natural populations of 328 mammalian species, accurately predicting cancer incidence rates with known rates for colorectal and lung cancers. Thesis available here.