The finding that interdisciplinary research has low funding success touches a sore spot in molecular biology (see L. Bromham et al. Nature 534, 684–687; 2016). The skilful integration of physics, mathematics and biology that led to the development of molecular biology is being superseded by the use of bioinformatics tools that can process and visualize large amounts of experimental data. Yet these tools often deliver only incremental advances in complex topics (for instance, in the function of transcriptional networks).

Genuinely interdisciplinary landmark discoveries include the stochastic nature of gene expression and the realization that biological systems are 'noisy' (M. B. Elowitz et al. Science 297, 1183–1186; 2002), and the finding that there was interbreeding between Neanderthals and ancestors of modern humans (R. E. Green et al. Science 328, 710–722; 2010. That discovery relied on sophisticated sample-preparation methods and advanced statistical analysis to reconstruct the flow of genetic material between ancient genomes.

Historically, scientific curiosity has been driven by interdisciplinary knowledge. Gregor Mendel, for example, trained as a physicist. Modern teaching tends to gloss over the mathematical insights that his theory of inheritance required. I suspect that few biologists today could identify binomial distributions in pea-plant cross-breeding experiments and conclude that independent alleles are randomly segregated.