Table 5. GLM analysis of change in PLT counts shows best model is useful (p = .0033, R2 = 16.7%).

Source	Df	Sum of Squares	Mean Square	F value	Pr>F
Model	14	54700.2	3907.2	2.47	0.0033
Error	173	273894.6	1583.2
Corrected Total	187	328594.9
R_square	Coeff Var		Root MSE	PLT Change Mean
0.167	988.168		39.79	4.0266
Source	Df	Type III SS	Mean Square	F-value	Pr>F
ITPase	1	266.33	266.33	0.17	0.6822
Arm	1	5192.79	5192.79	3.28	0.0720
ITPase*Arm	1	7353.25	7353.25	4.64	0.0325
rs12979860	2	2485.27	1242.64	0.78	0.4578
ITPase* rs12979860	2	2521.016	1260.51	0.80	0.4527
Arm* rs12979860	2	2896.34	1448.17	0.91	0.4026
Age	1	638.71	638.71	0.4	0.5262
ITPase*Age	1	724.56	724.56	0.46	0.4996
Age*Arm	1	65.33	65.33	0.04	0.8393
Sex	1	5235.88	5235.88	3.31	0.0710
Arm*Sex	1	442.42	442.42	0.28	0.5977