What was measured in the embryo?
How sex is determined in turtles varies from species to species. In Chrysemys picta the temperature of the nest decides; in Apalone spinifera the chromosomes do. Setting the two side by side is the direct way of asking what changed in the transition between them. The team sampled urogonadal development in both species across five matched stages, incubating the eggs at 26 degrees Celsius, which yields 100 per cent males in Chrysemys, and at 31 degrees Celsius, which yields 100 per cent females.[1]
It matters what the scale actually is. Two biological replicates were generated for each condition, and each RNA library held RNA pooled from 11 to 15 embryos; at least 40 million clean 150 base pair paired-end reads were produced per library. The read depth is high and the number of biological replicates is low. A design like this gives the average expression pattern across pooled tissue rather than what happens in any individual embryo.[1]
How the cilia candidate emerged
The networks themselves are inferred rather than measured. The team combined gene expression, transcription factor binding sites and protein-protein interaction data using a message-passing method called PANDA, then tested four alternative hypotheses: are the hubs and their targets conserved, did the same hub acquire new targets, did a new hub take over the old targets, or did both change? What PANDA produces is a consensus map built from three datasets.[1]
The route to cilia is indirect as well. The known thermosensitive components overrepresented in the results, namely calcium-redox, pSTAT3, Wnt/Rspo1/beta-catenin and Dhh, are linked to primary cilia; and the transcription factors that evolved between the two species also regulate cilia, from different directions: calcium and ion channels and membrane transport components in Chrysemys, structural elements and ciliogenesis in Apalone. One reading is that cilia really are the structure gathering the temperature signal. There is another: primary cilia are annotated as connected to a great many developmental pathways, so the overrepresentation may partly reflect the general connectivity of the annotation databases used.[1]
The naming is itself a result
The authors do not hide that uncertainty. They call the finding the Primary Cilia Integration hypothesis and write plainly that it warrants functional validation. They also state themselves that the similarity among networks at the global level is likely to come from the small size of the dataset and the pooling of five developmental stages, and therefore treat what they have as conservative, meaning only the strongest signals could show through.[1]
Naming a hypothesis is no small act in biology. Temperature-dependent sex determination has been known for decades, but the question of which cellular structure the embryo senses temperature with stayed open. What this study delivers is a target with a name: from here on, ciliary function can be disrupted directly and the sex ratio can be checked for a shift. The mechanism itself remains undemonstrated, and what serves as the thermometer inside a turtle embryo is still unknown.[1]