3 ms·
They chose to show the cortisol data from the last paper[1] as a schematic...strange choice. Compare Fig 1F in this new one to Fig 1 from the previous work. The
by nonbel 11y ago
They chose to show the cortisol data from the last paper[1] as a schematic...strange choice. Compare Fig 1F in this new one to Fig 1 from the previous work. The unexplained difference between control groups from the two papers is bigger than the effect they see due to injecting micro RNAs.
[1] http://www.jneurosci.org/content/33/21/9003.abstract http://www.jneurosci.org/content/33/21/9003.abstract
- nonbel 11y agoI extracted the data from the two figures and made a plot: http://s21.postimg.org/8it9wt9ut/Cort.png http://s21.postimg.org/8it9wt9ut/Cort.png At least the 2013 Control and 2015 PBS (the control for the current study) should be similar. They need to explain why the results are so different and figure out what is required to get reproducible results before they conclude the treatment is responsible here. Also, in the supplemental methods they write: "Data greater than 2 SDs above each group mean were removed from analyses". That is not a legitimate reason to drop animals from the study. It appears they dropped four mice only to create the impression of less variable outcomes (and to help get "significance"). I am not confident this effect would replicate in an independent lab, they seem to be playing games with the analysis.