Were you lied to in your genetics class?

There's a disturbing trend in popular science these days. The goal is to convince the general public that much of what we thought we knew is wrong. I think it's related to the general mistrust of science.

A recently posted YouTube video tries to make the case that you were lied to about genetics. I'll get to than in a little while but first let me summarize what I was taught in a university genetics class in 1965.

Like many of you, I was taught the rudiments of Mendelian genetics in high school. This view was emphasized in my university course where we learned about Mendel's laws of inheritance in diploid species. I also learned about population genetics and exceptions to the standard rules governing the segregation of alleles. We covered 'phage and bacterial genetics where Mendel's laws mostly don't apply and we learned about sex-linked alleles that show more complicated inheritance.

I learned about recombination and linkage and I have fond memories of the lab part of the course where we did experiments with bacteria, corn, fruit flies, and Neurospora—all of which illustrated the basic principles of genetics. I liked the course so much that I applied for and was accepted for a summer job with my genetics professor (George Setterfield)—this turned out to be the first step in my career as a scientist.

There are hundreds of excellent examples showing a direct link between genotype and phenotype but there are also many examples where this connection is obscure, especially in complex multicellular organisms such as fruit flies, plants, mice, and humans. For example, there's no obvious reason why a deficiency in a starch branching enzyme should cause wrinked peas [Biochemist Gregor Mendel Studied Starch Synthesis]. And there's no obvious reason why mutations in the same gene in humans causes a liver disease in children that's usually fatal (Anderson disease) [Glycogen Storage Diseases]. The mutant alleles still segregate according to Mendelian genetics.

Genes interact to form networks and some of these networks are complicated. We usually teach simple examples such as how suppressor genes can mitigate the effect of stop mutations in bacteria. There are more complicated examples such as the large number of alleles and different genes that affect eye color in humans [The Genetics of Eye Color]. When the number of different loci is low, you can still work out the Mendelian probabilities using Punnett Squares but in some cases there are just too many loci with low penetrance.

This brings me to the video post on the SubAnima web site. The author is Jake Brown, a master's student in mathematical biology at the University of Melbourne (Australia). He has many interests but the ones most relevant to this topic are evolutionary biology and philosophy of biology. Jake thinks that you have been lied to about genetics because most of what you learned about Mendelian genetics is wrong.

In order to make his case, he focuses on the complicated examples of multigene inheritance in humans. The interesting part of the video is his version of Waddington's genetic landscape and his toy model illustrating the concept. I think he's dead wrong about the death of Mendelian genetics but it's still worth watching the video.

---