Biometricians vs Mendelians: What a century old debate ago can teach researchers today

by Sean Bankier, The University of Edinburgh

There is a historical precedent for major scientific advances that have come from those who have not necessarily followed the traditional academic route for their field. The field of genetics has been no exception to this phenomenon, from its conception by a particularly tenacious Augustinian monk to the somewhat less well-known story of a certain English public school teacher, Ronald Fisher.

In October 2018 I, with a group of geneticists, statisticians and mathematicians came together to celebrate the work of Ronald Fisher and his contribution to the field of quantitative genetics. Quantitative genetics is used to examine the genetic variation responsible for the emergence of complex traits, often involving multiple genes. This typically involves the use of different statistical and mathematical methods to examine continuous traits such as height, where the genetic variation responsible can be quantified.

Around the time of this meeting, 100 years ago, Fisher was publishing his seminal paper “The Correlation between Relatives on the Supposition of Mendelian Inheritance” that effectively founded the field of quantitative genetics. This meeting a century later, took place at the Royal College of Surgeons in Edinburgh where Fisher had originally published his manuscript. Quantitative genetics has come a long way in 100 years so this was a unique opportunity to reflect upon the state of the field and its many applications, from plant breeding, to the genomic architecture of complex traits and even the genetics of the human face.

Now celebrated as a major turning point for what we now define as complex trait genetics, at the time it didn’t quite have the impact that Fisher had hoped for, in no small part due to the fact that many of the geneticists at the time couldn’t understand it! Notoriously opaque and filled with equations and statistical jargon, this proved a challenge for many of the more experimentally minded researchers of the time. However, it was through this paper that Fisher was eventually able to reconcile the views of both sides of the raging debate between the “Biometricians” and the “Mendelians” (Figure 1).

Sean 2

Figure 1: From left to right. William Bateson, one of the strongest proponents of the Mendelian model in the early 20th century. Ronald Fisher, statistician and geneticist. Francis Galton, statistician responsible for the concepts of correlation and regression to the mean.

Charles Darwin’s views on evolution and inheritance had gained traction within certain factions of the scientific community, but Darwin himself was unable to propose an exact mechanism of how this inheritance was brought forward from one generation to the next. As the name would suggest, the Mendelians, were focused upon the model of inheritance that Mendel had put forward and like Mendel, these researchers tended to hail from an experimentalist background. Biometricians on the other hand, were influenced by ideas from Francis Galton and believed that the complexity of genetic variation could only be accounted for through the application of quantitative methodologies to fulfil the grand ideas which Darwin himself had set forth.

As the name of Fisher’s paper would suggest, he managed to integrate the key concepts of Mendelian Inheritance within a quantitative framework through utilisation of many of the statistical methods such as correlation that had been proposed by Galton himself. Although a challenge to read, the message behind Fisher’s paper is quite straight forward and demonstrates that Mendelian inheritance was best positioned to explain the continuous traits that the Biometricians tended to focus upon.

Apart from as an interesting side-note in history, there is a greater message that can be taken from this century old debate. Factionalism still exists today within science and the divisions that are drawn are not so different than they were in Fisher’s day. With the advent of high throughput genome sequencing technologies and large multi-omic datasets, there is as great a need for quantitative methods to analyse these data than there ever has been. However, as many are keen to point out, mathematical and statistical models developed to explain biology must be grounded within a biological context. This is true, given the well-worn adage that correlation does not always equal causation (Figure 2).

sean 3

Figure 2: Correlation does not always equal causation… or does it? (r=0.66) Image taken under creative commons license with thanks to Tyler Vigen

The use of both quantitative and experimental methodologies is crucial to tackle the major issues in genetics such as understanding complex disease, accounting for missing heritability and the integration of multiple layers of genetic information. Collaboration with mathematicians to provide models in systems biology or physicists to understand chromatin dynamics and proteins folding will be essential in advancing the state of the field and should be encouraged. The challenge is more human than scientific, as what is required is communication between disciplines. In addition to my own PhD, I help organise a monthly meet up group to discuss synthetic biology with individuals across a wide range of disciplines and backgrounds. This experience has taught me that lack of collaboration with other fields is not down to absence of interest but want of opportunity.

As researchers within genetics come together from increasingly specialised fields, each with their own pre-conceptions and biases, perhaps what is needed is a 21st Century Ronald Fisher to bring together the disparate ideas across biology to address these larger questions. As cross-disciplinary collaborations become more popular and incentives for such projects more common, the future for another 100 years of quantitative genetics becomes increasingly exciting and vibrant.

Further reading:

Fisher, R.A. (1919). XV.—The Correlation between Relatives on the Supposition of Mendelian Inheritance. Earth Environ. Sci. Trans. R. Soc. Edinb. 52, 399–433.

Knapp, B., Bardenet, R., Bernabeu, M.O., Bordas, R., Bruna, M., Calderhead, B., Cooper, J., Fletcher, A.G., Groen, D., Kuijper, B., et al. (2015). Ten Simple Rules for a Successful Cross-Disciplinary Collaboration. PLoS Comput. Biol. 11.

Visscher, P.M., and Bruce Walsh, J. Commentary: Fisher 1918: the foundation of the genetics and analysis of complex traits. Int. J. Epidemiol.

About Me:

sean 1Sean is an PhD student using systems genetics to identify causal gene networks in complex disease, with a focus upon plasma cortisol and cardiovascular disease at the Institute of Bioengineering of the University of Edinburgh. Sean is also a co-organiser of Cafe Synthetique Edinburgh (@CafeSynthEdin), a group that runs a series of informal events aimed at starting conversations around synthetic biology that are accessible for everyone, regardless of background.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s