This research has a very small sample size (n=31)
Also indicated by the scatter plots too.
From the research: “total of 10 toddlers with ASD and 6 controls ... In a 2021 batch, we measured BCOs from 10 ASD and 5 controls.“
https://www.researchgate.net/publication/380876463_Embryonic...
I haven't read the paper yet, but if its like what our team is doing, we take blood, and turn the cells there into neurons, then organoids.
however, then correlating the consistent organoid attributes back to attributes of a small parent sample group is weaker on small sample groups than larger ones. here the parent group is tiny… and the severe ASD sufferers is even smaller (n=2 or 3 by my understanding)
great foundational research! but surely larger group sizes are required before correlation is confirmed