ERC Starting Grant founding!

How it happened…

In 2022, I submitted an ERC Starting Grant “EvoGenArch” on the study of the distribution of the effects sizes of variants influencing phenotypes in the genomes (more details below). In the first round, I was invited to the second step of the process (oral presentation of the project), but didn’t make the principal or complementary list… Well, that happens, right? But as it turned out, due to a shift in the calendar, I had to submit my project again (during Autumn 2022), without having access to the panel and reviewers’ output of my first trial… I wasn’t sure this second attempt would be any better. But… I did made it to the complementary list this time! The ERC warned me (in Autumn 2023 then) that they could call projects down the list up until December. After that, nope!

Time passed, including December and I was relatively sure I didn’t get the founding (I didn’t apply again, because I wanted time to consolidate further the application for a Consolidator grant). But in March 2024 (this is a long process isn’t it?), I was extremely surprised to get an email from the European Commission titled “Start of grant preparation” stating that my proposal “has reached the stage of Grant Agreement preparation”. I first thought about a technical error, given that I received absolutely no notice of getting the grant, and it was already way too late according to the ERC’s own timeline. But it turned out to be legit (notably, I found a proper letter about it by rummaging through their online platform), I got the ERC Starting Grant!

Dance of joy!

What is it about: the genetic architecture problem

When Darwin published On the Origins of Species in 1859, he was missing a particularly important piece of the puzzle: a proper theory of heredity (well, he had a theory called pangenesis, but let’s just say that this one didn’t stick as much as the rest of his work did…). This didn’t stop Darwin (for very good reasons) to conceive the random modifications in the heredity process (what we now call “mutations”) as a gradual, smooth process. When the laws of Mendel were rediscovered near 1900 and genetics arose as the theory of heredity, the view that those random modifications were actually discrete “jumps” of medium to large importance (hence the word “mutations” we are still using) became more in line with Mendelian inheritance and gained quite a lot of traction.

Size is a quantitative trait

But think about it: how to reconcile Mendelian inheritance and this mutationist view of medium to large modifications inherited through the segregation of alleles, with a trait like… size? We all observe that size is rather inherited as a “gradual” thing (a quantitative trait) rather than in lumps of “large” or “small” alleles. This simple fact actually generated quite some friction in the full adoption of the newborn genetics as a universal mechanism for heredity. This problem was solved by a famous paper from Fisher in 1918, demonstrating that we could model quantitative traits by assuming a very large number of genes, with Mendelian inheritance, participate to its variation among individuals.

But then, the mutations for those quantitative traits, are they rather small, corresponding to the gradual view of Darwin, or are at least some them rather large, more in line with the views of the first geneticists? Fast forward to today, and… it is still very hard to answer that question! In fact, there is still some controversy to it: a lot of theoretical arguments were advanced by Fisher, Kimura and Orr about whether we should expect small, medium or a mix of small/large effects in the mutations affecting quantitative traits (we call them QTL, for Quantitative Trait Loci), but they are based on very general models and as soon as we modify some key assumptions, we see the predictions of the model to diverge from those general views. This means that biologists are still battling quite fiercely about whether the “stuff of evolution” is made of smooth powder or big rocky chunks! The expression “genetic architecture” is used to refer at how the genetic basis of quantitative traits is structured, and can mean at lot of different things. Amongst others, it means how small/large are the effects of QTL, and this is strictly how “genetic architecture” should be understood here (yeah, I know, there’s a lot more to it than this sole problem, but we have to start somewhere, don’t we?).

A rare footage of evolutionary biologists arguing about the effect sizes of QTL

One key aspect of the small/large effects debate is that it has been mainly focused on providing a “catch-all” theory of all quantitative traits. But if the theoretical models have taught us something, it is that whether we should expect rather smaller or larger effects for QTL largely depends on the evolutionary context within which the trait itself evolved. So, my ERC project is about the following: how to properly study the genetic architecture (distribution of the effects of QTL) of quantitative traits in wild populations, and relate the variation that we might observe in genetic architecture to key evolutionary parameters (type of selection, historical demography, local adaptation…)?

The EvoGenArch ERC project

The project has three main components (this is were it gets a wee bit more technical…):

  1. Develop a method that is focused on the inference of the distribution of the effects of QTL (“genetic architecture”). Most existing methods assume that the QTL are among the available genetic markers (genetic variations in the genotype that we use as data in the statistical methods), which is problematic for a few reasons (mainly, because QTL can be any kind of mutation, not necessarily the one kept in the set of genetic markers).
  2. Use the above method in the context of our common lizard study, after having sequenced many individuals from (1) our wild population surveyed in the Cevennes and (2) an ongoing “common garden” experiment (where individuals from different populations are raised in a common environment). The idea is to be able to link the genetic architecture of various traits and how selection affects them in the wild, and signals of local adaptation as inferred from the common garden experiment.
  3. Use the method on a diversity of species (right now more than 20 species, but this is meant to grow) to compare the genetic architecture of comparable traits across species, accounting for cursory information regarding the historical and evolutionary context of each species/population and the selection acting the traits.

The idea is that, by the end of the project, we have a better understanding of the variety of the genetic architectures that we should expect out there, and refine our expectations according to some cursory knowledge about the evolutionary context. This is important and will have consequences throughout evolutionary biology and genetics!

Hiring incoming!

Keep an eye out for job proposals in the very close future (one post-doc and one research engineer to come for hiring in September).