Read on Twitter
Observation. Since moving into the neurological disease field, I've found that many (reviewers!) feel that animal models need to precisely phenocopy human symptoms to be 'valid'. A brief thread on why this view may limit important contributions from model organisms.
Despite conservation of genes and genetic networks, genotype-phenotype relationships can still vary drammatically between divergent species (such as humans and Drosophila). I often re-read this inspiring review by @BenLehner that expands on this point. https://www.nature.com/articles/nrg3404?message=remove&lang=en?WT.ec_id=NRG-201303
Such variances reflect the obvious fact that these species possess radically different body plans that are adapted to distinct ecological niches. In other words, developmental and evolutionary context of gene network activity is critical.
Fundamentally, when we use flies to model disease, the questions I am interested in are: what does a disease-linked gene do in relevant cell-types? What genetic networks does it act in? Can we identify genetic suppressors that might suggests new therapies?
To answer these questions we may need an organismal phenotype as a readout, but we shouldn't expect that it will mimic the phenotype caused by mutations in the human orthologue. Wangler, Bellen and colleagues have put it much better than I. https://www.genetics.org/content/207/1/9.long
This logic also applies to vertebrate models. Mutations in a gene of a nocturnal tetrapod that shared a common ancestor with humans >65 MY ago (a mouse) may well yield distinct phenotypes compared to the orthologous mutation in a diurnal, bipedal species (humans).
Let's accept biological reality and embrace diversity in genotype-phenotype relationships. Every tool we possess needs to be deployed to study disease. Worms, flies, yeast etc. should not be discounted: all can provide important biological insights. https://science.sciencemag.org/content/302/5651/1772.long
Now, back to my Fellowship application on Drosophila disease models....
