fragile reversal, page-3

I have to say that I'm a little unconvinced by the mouse data. This is why:

1. The experiments involve treatment of wildtype mice and gene-knockout mice deemed to be a model of human fragile X syndrome. Both groups are treated with NNZ-2566 or a control. The one comparison you'd want to see is the impact of NNZ-2566 on the KO mice compared to the KO mice that did not get the treatment. Despite this, statistical comparisons are made between each other pair of groups EXCEPT that one. Why not?

2. n values (number of mice) and number of experiments is not shown. Why not? We want to see lot sof mice and lots of repeat experiments. When people don't show this, it may be because they've tested very few mice.

3. look at the first graph (open field rearing). It looks a big difference, but now look a thte scale. instead of it being a 100% axis, its a 35% max, designed to make diffs look bigger. No statistical significance tested. Why not?

4. Look at the second graph (learning). A statistically signif diff is shown between treated WT and treated KO mice. So what? That test says nothing about the KO mice other than than respond to the drug better than the WT mice.The real thing you want to see is the KO mice responding to the drug (ie KO control versus KO treated). Why is this not tested?

5. Look at the same graph (learning). Look at the comparison between the vehicle treated mice. The testing says 'ns' ie statistically identical. This sis worry because the KO mice are supposed to reperesent the disease, yet their own test says there is no difference between them. Now look at the comparison between the treated mice (green columns). Depsite a much smaller difference between teh graphs, this apparantly IS signif. Suggests weird stats or very weird things with the number of mice.

6. look at graph 3. (social behaviour).
Again, the crucial test of checking effect of nn-2566 on KO mice is not statistically tested. Despite the formattingh errors re labelling of x-axis, the only signif diff they see is between KO and WT mice. Nothing else of significance shown. Null result.

Happy to hear counter views on this.