The shifting emphasis in genetics from single genes to gene interactions and networks was well illustrated at the recent 'Quantitative Genetics and Genomics' Gordon Conference. The organizers, Steve Knapp and Bruce Walsh, brought together researchers with wide-ranging interests — from statistical and population genetics, to human disease and plant breeding. Participants were left in no doubt that understanding the genetic basis of polygenic traits is a realizable goal.

Some of the challenges that face studies into polygenic disease are illustrated in our lead review this month, although the focus is a 'classic monogenic' human disorder — β-thalassaemia. To study the striking phenotypic variation in this disorder, David Weatherall emphasizes the importance of understanding disease pathophysiology. The work also shows how the distinction between polygenic and monogenic disorders is beginning to blur. Another challenge for polygenic disease research concerns the requirement for huge sample sizes. Practical difficulties aside, sample collections and databases on this scale raise new concerns about privacy and potential harms to individuals. Ruth Chadwick and Kåre Berg provide an ethical perspective on ways to address these concerns.

The most complete genetic inventories for specific processes or phenotypes are available in model systems — innate immunity in Drosophila, as reviewed by Deborah Kimbrell and Bruce Beutler, is a case in point. However, even when a process can be described in exquisite biochemical detail, such as for small, gene regulatory networks in Escherichia coli, its description is still incomplete. Jim Collins and colleagues show how computational modelling can help. Models of natural and synthetic prokaryotic genetic networks can complement experimental work, and indicate new avenues for research — a glimpse of what's to come for higher systems.