Quantitative genetics, the analysis of the genetics of complex quantitatively varying traits, primarily through statistical approaches, has played a significant role in the twentieth and twenty-first century life sciences but has received less critical attention than other areas of genetics such as Mendelian genetics, molecular biology and genomics. Recent years have seen a move to address this gap in historical, philosophical and social studies (e.g. Serpico et al. 2023): Alongside longstanding research on the Biometrician-Mendelian debate out of which quantitative genetics first developed, there has been significant attention given to the use of quantitative genetics in the controversial area of human behavioural genetics (Panofsky 2014; Kaplan & Turkheimer 2021), and more recently to the application of quantitative genetic techniques to disease risk prediction in biomedicine (Serpico 2023). Separately, researchers in the history, sociology and philosophy of agriculture have contributed to our understanding of a different lineage of quantitative genetic practices in plant and animal breeding (Theunissen 2020; Derry 2020; Lowe and Bruce 2019; Williamson and Leonelli 2021). Despite significant overlap between the methods and approaches of quantitative genetics in biomedicine and behavioural research on the one hand and agriculture on the other, there has been little cross-discussion between scholars working on these distinct domains. This roundtable discussion aims to open an interdisciplinary dialogue on quantitative genetic practices across domains and species, through contributions from philosophers and social scientists, addressing commonalities and discontinuities in the epistemology, politics and ethics of quantitative genetic approaches. In particular, the session will focus on the intersection of statistical and genomic approaches and their implications, for example in predictive practices such as Polygenic Risk Scores for human disease and Genomic Selection in plant and animal breeding. Participants will provide short introductions to key cases before opening the discussion to the panel and wider audience.
Quantitative genetics has received less attention from philosophers, historians and social scientists than other areas of genetics such as Mendelian genetics, molecular biology and genomics. In recent years there has been a turn on the part of historians and philosophers to analyse the uses and implications of quantitative genetics in the controversial area of human behavioural genetics, and increasingly in areas of biomedicine. Separately, there is a small body of work on quantitative genetic practices in agricultural domains, especially plant and animal breeding, largely conducted by social scientists and often situating these in the political economic contexts of farming. These two strands of work have not yet been brought into dialogue with one another to identify lines of comparison in how genetic phenomena are understood and manipulated, how quantitative genetic work is organized, and what political and ethical implications arise from the different applications of quantitative genetics. This talk introduces the session and presents a perspective on quantitative genetics in plant and animal breeding. In particular, I analyse how this domain is organized around the production, circulation and use of indicators—both statistical (such as heritability, breeding values, and genetic gain) and material (such as genome-wide SNP marker panels)—and has been marked by exchanges of repertoires between the plant and animal domains.
Genetic prediction using quantitative genetics methods has long been applied in agriculture and animal breeding. However, the reliability and clinical utility of such methods for predicting human complex diseases remain controversial. This talk examines the challenges in leveraging genome-wide association studies (GWAS) and polygenic scores (PGSs) to predict human phenotypes, especially in light of the dynamic nature of gene expression and the outstanding variability between individuals.
I analyse some key factors that may limit the predictive capability of PGSs for human complex traits, including the relationship between genetic penetrance, phenotypic plasticity, and individual variability in biological, environmental, and developmental factors. Many human traits under the lens of human medicine and behavioral genetics, such as intelligence, personality, and mental disorders, are highly plastic and sensitive to environmental influences.
As a result, the underlying genes may exhibit reduced penetrance or significant variability in expression across individuals and developmental stages. For example, the predictive power of PGSs for educational attainment is modest, as this trait is heavily influenced by socioeconomic status, quality of education, and other environmental factors. Similarly, the expression of genes associated with depression can vary greatly depending on individual experiences of adverse childhood events and psychosocial distress.
In conclusion, the limited predictive ability of PGSs for human diseases may reflect differences between human phenotypic development and other domains where genetic prediction has been more successful. Accounting for trait canalization, plasticity, and individual variability is crucial for enhancing the reliability and clinical utility of genomics-based approaches in medicine.