Single-cell experiments that invite statistical physics in genetics

par Gaël Yvert

vendredi 4 avr. 2014, 14:00 → 15:00 Europe/Paris

Auditorium M. Vivargent (LAPP)

Genetics aims at discovering how the genotype of an organism (its DNA sequence) modifies its phenotype (quantitative or qualitative values of physiological traits). This has long been pursued from a very deterministic angle: DNA mutations that were found are those that modify the average trait values of individuals carrying them. Although powerful in many cases, this approach has important limitations. It does not explain why a given mutation can have a dramatic effect in an individual and a mild one or no effect in others. It also fails to detect many mutations that contribute little to the trait. Current genetics therefore faces a double challenge: understanding the causalities that were found and finding the hidden ones. As an alternative approach, we are exploring genetic control from a probabilistic angle. By studying the statistical distribution of biological traits among individual cells (the 'particles' of biology), it is possible to observe how the genotype changes the probability law of a phenotypic outcome. We initiated this approach on yeast cells, where we found three mutations that alter variance in gene expression. We also observed that natural yeast genotypes confer variable levels of "noise" in cellular morphology. Such probabilistic effects of DNA mutations invite us to study organisms as a statistical system of cells and to discuss genetic predisposition with physicists.

Slides 2014-04-04_genetics.pdf