. 2020 Mar 3;7(2):ENEURO.0285-19.2020. doi: 10.1523/ENEURO.0285-19.2020

Table 1.

Statistical tests used to analyze data

Feature	Comparison	Data structure (D’Agostino normality test)	Type of test	Sample size (# of NMJs, # of animals)	Mean	Statistic	p Value	Outlier test
Structure
Bouton number	w¹¹¹⁸ vs cow^GDP/Df	Normal	Unpaired two-tailed t test	w¹¹¹⁸ (15,8); cow null (16,8)	23.53 vs 41.13	t = 8.296 df = 29	p < 0.0001
Bouton number	vglut-GAL4 vs UAS-cow-RNAi	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	vglut-GAL4 (16,8); UAS-Cow-RNAi (16,8)	26.69 vs 28.8	Mean rank diff = −2.938	p > 0.9999
	vglut-GAL4 vs vglut>cow-RNAi			vglut-GAL4 (16,8); vglut>cow-RNAi (15,8)	26.69 vs 37.38	Mean rank diff = −19.09	p = 0.0002
	UAS-cow-RNAi vs vglut>cow-RNAi			UAS-cow-RNAi (16,8); vglut>cow-RNAi (15,8)	28.8 vs 37.38	Mean rank diff = −16.16	p = 0.0031
Bouton number	24B-GAL4 vs UAS-cow-RNAi	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	24B-GAL4 (16,8); UAS-cow-RNAi (16,8)	30.63 vs 31.5	Mean rank diff = −6.188	p = 0.6307
	24B-GAL4 vs 24B>cow-RNAi			24B-GAL4 (16,8); 24B>cow-RNAi (16,8)	30.63 vs 28.06	Mean rank diff = 3.563	p > 0.9999
	UAS-cow-RNAi vs 24B>cow-RNAi			UAS-cow-RNAi (16,8); 24B>cow-RNAi (16,8)	31.5 vs 28.06	Mean rank diff = 9.75	p = 0.1451
Bouton number	vglut/+ vs vglut>Cow	Normal	Unpaired two-tailed t test	vglut/+ (16,8); vglut>Cow (16,8)	25.25 vs 27.06	t = 1.122 df = 30	p = 0.2706
Bouton number	24B/+ vs 24B>Cow	Normal	Unpaired two-tailed t test	24B/+ (16,8); 24B>Cow (16,8)	30.38 vs 29.81	t = 0.2317 df = 30	p = 0.8183
Bouton number	FRT-Wg vs FRT-Wg;Cow^GDP	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	FRT-Wg (24,12); FRT-Wg;Cow^GDP (24,12)	26.71 vs 31.71	Mean rank diff = −22.29	p = 0.0300
	FRT-Wg vs NRT-Wg			FRT-Wg (24,12); NRT-Wg (24,12)	26.71 vs 27.04	Mean rank diff = −3.521	p > 0.9999
	FRT-Wg vs NRT-Wg;Cow^GDP			FRT-Wg (24,12); NRT-Wg;Cow^GDP (23,12)	26.71 vs 26.78	Mean rank diff = 0.4312	p > 0.9999
	FRT-Wg;Cow^GDP vs NRT-Wg			FRT-Wg;Cow^GDP (24,12); NRT-Wg (24,12)	31.71 vs 27.04	Mean rank diff = 18.77	p = 0.1085
	FRT-Wg;Cow^GDP vs NRT-Wg;Cow^GDP			FRT-Wg;Cow^GDP (24,12); NRT-Wg;Cow^GDP (23,12)	31.71 vs 26.78	Mean rank diff = 22.72	p = 0.0278
	NRT-Wg vs NRT-Wg;Cow^GDP			NRT-Wg (24,12); NRT-Wg;Cow^GDP (23,12)	27.04 vs 26.78	Mean rank diff = 3.952	p > 0.9999
Bouton number	w¹¹¹⁸ vs cow^GDP/+	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	w¹¹¹⁸ (15,8); cow^GDP/+ (15,8)	28.33 vs 35.73	Mean rank diff = −15.93	p = 0.0929
	w¹¹¹⁸ vs Notum^KO/+			w¹¹¹⁸ (15,8); Notum^KO/+ (16,8)	28.33 vs30.75	Mean rank diff = −5.565	p > 0.9999
	w¹¹¹⁸ vs. cow^GDP/Notum^KO			w¹¹¹⁸ (15,8); cow^GDP/Notum^KO (16,8)	28.33 vs 46.13	Mean rank diff = −35.81	p < 0.0001
	cow^GDP/+ vs Notum^KO/+			cow^GDP/+ (15,8); Notum^KO/+ (16,8)	35.73 vs 30.75	Mean rank diff = 10.37	p = 0.6569
	cow^GDP/+ vs cow^GDP/Notum^KO			cow^GDP/+ (15,8); cow^GDP/Notum^KO (16,8)	35.75 vs 46.13	Mean rank diff = −19.88	p = 0.0129
	Notum^KO/+ vs cow^GDP/Notum^KO			Notum^KO/+ (16,8); cow^GDP/Notum^KO (16,8)	30.75 vs 46.13	Mean rank diff = −30.25	p < 0.0001
Bouton number	w¹¹¹⁸ vs cow^GDP/cow^GDP	Normal	Ordinary one-way ANOVA with Tukey's multiple-comparisons test	w¹¹¹⁸ (18,10); cow^GDP/cow^GDP (19,10)	22.94 vs 33.74	q = 9.731 df = 76	p < 0.0001
	w¹¹¹⁸ vs Notum^KO/Notum^KO			w¹¹¹⁸ (18,10); Notum^KO/Notum^KO (20,10)	22.94 vs 30.5	q = 6.897 df = 76	p < 0.0001
	w¹¹¹⁸ vs cow^GDP,Notum^KO/cow^GDP,Notum^KO			w¹¹¹⁸ (18,10); cow^GDP,Notum^KO/cow^GDP,Notum^KO (23,12)	22.94 vs 29.13	q = 5.83 df = 76	p = 0.0005
	cow^GDP/cow^GDP vs Notum^KO/Notum^KO			cow^GDP/cow^GDP (19,10); Notum^KO/Notum^KO (20,10)	33.74 vs 30.5	q = 2.996 df = 76	p = 0.1564
	cow^GDP/cow^GDP vs cow^GDP,Notum^KO/cow^GDP,Notum^KO			cow^GDP/cow^GDP (19,10); cow^GDP,Notum^KO/cow^GDP,Notum^KO (23,12)	33.74 vs 29.13	q = 4.407 df = 76	p = 0.0135
	Notum^KO/Notum^KO vs cow^GDP,Notum^KO/cow^GDP,Notum^KO			Notum^KO/Notum^KO (20,10); cow^GDP,Notum^KO/cow^GDP,Notum^KO (23,12)	30.5 vs 29.13	q = 1.328 df = 76	p = 0.7838
% Satellite Boutons	w¹¹¹⁸ vs cow^GDP/Df	Normal	Unpaired two-tailed t test	w¹¹¹⁸ (15,8); cow null (15,8)	3.301 vs 3.336%	t = 0.03021 df = 28	p = 0.9761	ROUT, Q = 1%, removed 1 cow^GDP/Df value
% Satellite boutons	vglut-GAL4/+ vs UAS-Cow-RNAi/+	Normal	Ordinary one-way ANOVA with Tukey's multiple-comparisons test	vglut-GAL4/+ (16,8); UAS-Cow-RNAi/+ (15,8)	2.895 vs 2.908%	q = 0.016 df = 42	p > 0.9999	ROUT, iQ = 1%, removed 2 vglut>Cow-RNAi values
	vglut-GAL4/+ vs vglut>Cow-RNAi			vglut-GAL4/+ (16,8); vglut>Cow-RNAi (14,8)	2.895 vs 5.772%	q = 3.309 df = 42	p = 0.0612
	UAS-Cow-RNAi/+ vs vglut>Cow-RNAi			UAS-Cow-RNAi/+ (15,8); vglut>Cow-RNAi (14,8)	2.908 vs 5.772%	q = 3.244 df = 42	p = 0.0677
% Satellite boutons	24B-GAL4/+ vs UAS-Cow-RNAi/+	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	24B-GAL4/+ (16,8); UAS-Cow-RNAi/+ (16,8)	0.88 vs 2.381%	Mean rank diff = −8.656	p = 0.1328
	24B-GAL4/+ vs 24B>cow-RNAi			24B-GAL4/+ (16,8); 24B>cow-RNAi (16,8)	0.88 vs 2.806%	Mean rank diff = −8.969	p = 0.1114
	UAS-Cow-RNAi/+ vs 24B>cow-RNAi			UAS-Cow-RNAi/+ (16,8): 24B>cow-RNAi (16,8)	2.381 vs 2.806%	Mean rank diff = −0.3125	p > 0.9999
% Satellite boutons	vglut/+ vs vglut>Cow	Not normal	Mann–Whitney test	vglut/+ (16,8); vglut>Cow (16,8)	2.326 vs 7.121%	U = 38	p = 0.0003
% Satellite boutons	24B/+ vs 24B>Cow	Normal	Unpaired two-tailed t test	24B/+ (16,8); 24B>Cow (16,8)	3.164 vs 5.476%	t = 1.177 df = 30	p = 0.2486
% Satellite boutons	FRT-Wg vs FRT-Wg;Cow^GDP	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	FRT-Wg (16,8); FRT-Wg;Cow^GDP (16,8)	2.038 vs 1.002%	Mean rank diff = 5.167	p > 0.9999
	FRT-Wg vs NRT-Wg			FRT-Wg (16,8); NRT-Wg (16,8)	2.038 vs 8.304%	Mean rank diff = −26.08	p = 0.0021
	FRT-Wg vs NRT-Wg;Cow^GDP			FRT-Wg (16,8); NRT-Wg;Cow^GDP (16,8)	2.038 vs 3.595%	Mean rank diff = −5.452	p > 0.9999
	FRT-Wg;Cow^GDP vs NRT-Wg			FRT-Wg;Cow^GDP (16,8); NRT-Wg (16,8)	1.002 vs 8.304%	Mean rank diff = −31.25	p = 0.0001
	FRT-Wg;Cow^GDP vs NRT-Wg;Cow^GDP			FRT-Wg;Cow^GDP (16,8); NRT-Wg;Cow^GDP (16,8)	1.002 vs3.595%	Mean rank diff = −10.62	p > 0.9999
	NRT-Wg vs NRT-Wg;Cow^GDP			NRT-Wg (16,8); NRT-Wg;Cow^GDP (16,8)	8.304 vs 3.595%	Mean rank diff = 20.63	p = 0.0038
% Satellite boutons	w¹¹¹⁸ vs cow^GDP/+	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	w¹¹¹⁸ (15,8); cow^GDP/+ (15,8)	1.89 vs 3.079%	Mean rank diff = −7.867	p > 0.9999
	w¹¹¹⁸ vs Notum^KO/+			w¹¹¹⁸ (15,8); Notum^KO/+ (16,8)	1.89 vs 3.379%	Mean rank diff = −10.95	p = 0.4586
	w¹¹¹⁸ vs cow^GDP/Notum^KO			w¹¹¹⁸ (15,8); cow^GDP/Notum^KO (16,8)	1.89 vs 3.337%	Mean rank diff = −13.2	p = 0.1961
	cow^GDP/+ vs Notum^KO/+			cow^GDP/+ (15,8); Notum^KO/+ (16,8)	3.079 vs 3.379%	Mean rank diff = −3.079	p > 0.9999
	cow^GDP/+ vs cow^GDP/Notum^KO			cow^GDP/+ (15,8); cow^GDP/Notum^KO (16,8)	3.079 vs 3.337%	Mean rank diff = −5.329	p > 0.9999
	Notum^KO/+ vs cow^GDP/Notum^KO			Notum^KO/+ (16,8); cow^GDP/Notum^KO (16,8)	3.379 vs 3.337%	Mean rank diff = −2.25	p > 0.9999
% Satellite boutons	w¹¹¹⁸ vs cow^GDP/cow^GDP	Not normal	Kruskal–Wallis with Dunn's multiple-comparisons test	w¹¹¹⁸ (18,10); cow^GDP/cow^GDP (19,10)	1.904 vs 1.623%	Mean rank diff = 2.006	p > 0.9999
	w¹¹¹⁸ vs Notum^KO/Notum^KO			w¹¹¹⁸ (18,10); Notum^KO/Notum^KO (20,10)	1.904 vs 2.443%	Mean rank diff = −1.989	p > 0.9999
	w¹¹¹⁸ vs cow^GDP,Notum^KO/cow^GDP,Notum^KO			w¹¹¹⁸ (18,10); cow^GDP,Notum^KO/cow^GDP,Notum^KO (23,12)	1.904 vs 0.5223%	Mean rank diff = 9.155	p = 0.7029
	cow^GDP/cow^GDP vs Notum^KO/Notum^KO			cow^GDP/cow^GDP (19,10); Notum^KO/Notum^KO (20,10)	1.623 vs 2.443%	Mean rank diff = −3.995	p > 0.9999
	cow^GDP/cow^GDP vs cow^GDP,Notum^KO/cow^GDP,Notum^KO			cow^GDP/cow^GDP (19,10); cow^GDP,Notum^KO/cow^GDP,Notum^KO (23,12)	1.623 vs 0.5223%	Mean rank diff = 7.149	p > 0.9999
	Notum^KO/Notum^KO vs cow^GDP,Notum^KO/cow^GDP,Notum^KO			Notum^KO/Notum^KO (20,10); cow^GDP,Notum^KO/cow^GDP,Notum^KO (23,12)	2.443 vs 0.5223%	Mean rank diff = 11.14	p = 0.2978
Expression
Cow intensity	vglut/+ vs vglut>Cow	Not normal	Mann–Whitney test	vglut/+ (16,8); vglut>Cow (16,8)	1 vs 3.035	U = 0	p < 0.0001
Cow intensity	24B/+ vs 24B>Cow	Not normal	Mann–Whitney test	24B/+ (16,8); 24B>Cow (16,8)	1 vs 3.907	U = 0	p < 0.0001
Wg intensity	vglut/+ vs vglut>Cow	Not normal	Mann–Whitney test	vglut/+ (16,8); vglut>Cow (16,8)	1 vs 0.6731	U = 46	p = 0.0014
Wg intensity	24B/+ vs 24B>Cow	Normal	Unpaired two-tailed t test	24B/+ (16,8); 24B>Cow (16,8)	1 vs 1.518	t = 3.266 df = 30	p = 0.0027
Wg intensity	w¹¹¹⁸ vs cow^GDP/+	Normal	Ordinary one-way ANOVA with Tukey's multiple-comparisons test	w¹¹¹⁸ (15,8); cow^GDP/+ (15,8)	1 vs 0.885	q = 1.328 df = 56	p = 0.7840
	w¹¹¹⁸ vs Notum^KO/+			w¹¹¹⁸ (15,8); Notum^KO/+ (15,8)	1 vs 1.095	q = 1.094 df = 56	p = 0.8660
	w¹¹¹⁸ vs cow^GDP/Notum^KO			w¹¹¹⁸ (15,8); cow^GDP/Notum^KO (15,8)	1 vs 0.9014	q = 1.139 df = 56	p = 0.8515
	cow^GDP/+ vs Notum^KO/+			cow^GDP/+ (15,8); Notum^KO/+ (15,8)	0.885 vs 1.095	q = 2.422 df = 56	p = 0.3268
	cow^GDP/+ vs cow^GDP/Notum^KO			cow^GDP/+ (15,8); cow^GDP/Notum^KO (15,8)	0.885 vs 0.9014	q = 0.1886 df = 56	p = 0.9991
	Notum^KO/+ vs cow^GDP/Notum^KO			Notum^KO/+ (15,8); cow^GDP/Notum^KO (15,8)	1.095 vs 0.9014	q = 2.234 df = 56	p = 0.3985
Brp punctae number	w¹¹¹⁸ vs cow^GDP	Normal	Unpaired two-tailed t test	w¹¹¹⁸ (15,8); cow^GDP (15,8)	193.1 vs 284.8	t = 6.152 df = 28	p < 0.0001
Brp punctae Volume	w¹¹¹⁸ vs cow^GDP	Normal	Unpaired two-tailed t test	w¹¹¹⁸ (15,8); cow^GDP (15,8)	0.8576 vs 0.7164 μm³	t = 3.429 df = 28	p = 0.0019	ROUT, Q = 1%, removed 1 cow^GDP value
Brp punctae number	w¹¹¹⁸ vs cow^GDP	Normal	Unpaired two-tailed t test	w¹¹¹⁸ (11,8); cow^GDP (10,8)	298.6 vs 387.9	t = 3.598 df = 19	p = 0.0019
GluR cluster number	w¹¹¹⁸ vs cow^GDP	Normal	Unpaired two-tailed t test	w¹¹¹⁸ (11,8); cow^GDP (9,6)	382 vs 542.8	t = 4.353 df = 18	p = 0.0004
Function
EJC amplitude	w¹¹¹⁸ vs cow^GDP	Normal	Ordinary one-way ANOVA with Tukey's multiple-comparisons test	w¹¹¹⁸ (26,20); cow^GDP (20,18)	171.6 vs 212.1 nA	q = 3.868 df = 53	p = 0.0227	ROUT, Q = 1%, removed 1 cow^GDP value
	w¹¹¹⁸ vs cow^GDP/Df			w¹¹¹⁸ (26,20); cow^GDP/Df (10,9)	171.6 vs 254.2 nA	q = 4.197 df = 53	p = 0.0123
	cow^GDP vs cow^GDP/Df			cow^GDP (20,18); cow^GDP/Df (10,9)	212.1 vs 254.2 nA	q = 1.063 df = 53	p = 0.7341
EJC amplitude	w¹¹¹⁸ vs cow^GDP/+	Normal	Ordinary one-way ANOVA with Tukey's multiple-comparisons test	w¹¹¹⁸ (10,6); cow^GDP/+ (11,6)	217.2 vs 234.9 nA	q = 0.9383 df = 40	p = 0.9101
	w¹¹¹⁸ vs notum^KO/+			w¹¹¹⁸ (10,6); notum^KO/+ (11,9)	217.2 vs 214.1 nA	q = 0.1649 df = 40	p = 0.9994
	w¹¹¹⁸ vs cow^GDP/notum^KO			w¹¹¹⁸ (10,6); cow^GDP/notum^KO (12,7)	217.2 vs 235.9 nA	q = 1.009 df = 40	p = 0.8911
	cow^GDP/+ vs notum^KO/+			cow^GDP/+ (11,6); notum^KO/+ (11,9)	234.9 vs 214.1 nA	q = 1.13 df = 40	p = 0.8543
	cow^GDP/+ vs cow^GDP/notum^KO			cow^GDP/+ (11,6); cow^GDP/notum^KO (12,7)	234.9 vs 235.9 nA	q = 0.05304 df = 40	p > 0.9999
	notum^KO/+ vs cow^GDP/notum^KO			notum^KO/+ (11,9); cow^GDP/notum^KO (12,7)	214.1 vs 235.9 nA	q = 1.208 df = 40	p = 0.8282
mEJC Frequency	w¹¹¹⁸ vs cow^GDP	Normal	Ordinary one-way ANOVA with Tukey's multiple-comparisons test	w¹¹¹⁸ (22,17); cow^GDP (21,15)	1.396 vs 1.765 Hz	q = 1.419 df = 53	p = 0.5780
	w¹¹¹⁸ vs cow^GDP/Df			w¹¹¹⁸ (22,17); cow^GDP/Df (13,11)	1.396 vs 2.41 Hz	q = 3.406 q = 53	p = 0.0503
	cow^GDP vs cow^GDP/Df			cow^GDP (21,15); cow^GDP/Df (13,11)	1.764 vs 2.41 Hz	q = 2.15 df = 53	p = 0.2897
mEJC Frequency	vglut-GAL4/+ vs vglut>Cow-RNAi	Normal	Unpaired two-tailed t test	vglut-GAL4/+ (10,7); vglut>Cow-RNAi (11,7)	1.497 vs 2.449 Hz	t = 2.142 df = 19	p = 0.0454	ROUT, Q = 1%, removed 1 vglut-GAL4/+ value
mEJC amplitude	w¹¹¹⁸ vs cow^GDP	Normal	Ordinary one-way ANOVA with Tukey's multiple-comparisons test	w¹¹¹⁸ (21,16); cow^GDP (21,15)	0.7518 vs 0.8682 nA	q = 2.506 df = 52	p = 0.1889	ROUT, Q = 1%, removed 1 w¹¹¹⁸ value
	w¹¹¹⁸ vs cow^GDP/Df			w¹¹¹⁸ (21,16); cow^GDP/Df (13,11)	0.7518 vs 0.7165 nA	q = 0.6647 df = 52	p = 0.8856
	cow^GDP vs cow^GDP/Df			cow^GDP (21,15); cow^GDP/Df (13,11)	0.8682 vs 0.7165 nA	q = 2.857 df = 52	p = 0.1175
mEJC amplitude	vglut-GAL4/+ vs vglut>Cow-RNAi	Normal	Unpaired two-tailed t test	vglut-GAL4/+ (11,7); vglut>Cow-RNAi (11,7)	0.8015 vs 0.8446 nA	t = 0.8011 df = 20	p = 0.4325
Frequency	vglut/+ vs vglut>RNAi	Not normal (Shapiro–Wilk normality test performed because N too small)	Mann–Whitney test	vglut/+ (7,4); vglut>RNAi (6,3)	1.617 vs 2.977 Hz/μm²	U = 7	p = 0.0513
Mean ΔF/F₀	vglut/+ vs vglut>RNAi	Normal (Shapiro–Wilk normality test performed because N too small)	Unpaired two-tailed t test	vglut/+ (8,4); vglut>RNAi (5,3)	0.7912 vs 1.058 ΔF/F₀	t = 3.013 df = 11	p = 0.0118