R: Dunnett test 方法

假設共有4組試驗, 其中F1(LA1312xLA1543)為對照組, 另有3 組實驗組(親本LA1312,LA1543,LA1589)。每組試驗各有5重複。

1.安裝package-> multcomp

#20個番茄果實長寬比的觀測值
>y=c(1.233754545,1.1738625,1.21531,1.3708,1.251022222,1.464914286,1.4413,1.266463636,1.492966667,1.30975,1.302936364,1.296875,1.2109,1.286966667,1.2878,1.013746667,0.986116667,1.008646154,1.0426,1.031922222)

#將其分為四組，共五重複
> trt=gl(4,5,label=c("F1","LA1312","LA1543","LA1589"))

#打開package，沒有打開就做不了，所以一定要用library打開。
> library(multcomp)
Loading required package: mvtnorm
Loading required package: survival
Loading required package: splines

#執行ANOVA，發現其有顯著差異，再利用Dunnett 去看是哪幾個跟對照組有差異。
> out=aov(y~trt)
> summary(out)
            Df Sum Sq Mean Sq F value   Pr(>F)    
trt          3 0.3764 0.12548   28.75 1.11e-06 ***
Residuals   16 0.0698 0.00436                     
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

#glht 來做單一對比與直交對比
> glht(out,linfct=mcp(trt="Dunnett"))

         General Linear Hypotheses


Multiple Comparisons of Means: Dunnett Contrasts




Linear Hypotheses:
                 Estimate
LA1312 - F1 == 0  0.14613
LA1543 - F1 == 0  0.02815
LA1589 - F1 == 0 -0.23234

> dunnett.out=glht(out,linfct=mcp(trt="Dunnett"))
> summary(dunnett.out)

         Simultaneous Tests for General Linear Hypotheses


Multiple Comparisons of Means: Dunnett Contrasts


Fit: aov(formula = y ~ trt)


Linear Hypotheses:
                 Estimate Std. Error t value Pr(>|t|)    
LA1312 - F1 == 0  0.14613    0.04178   3.497  0.00837 ** 
LA1543 - F1 == 0  0.02815    0.04178   0.674  0.84038    
LA1589 - F1 == 0 -0.23234    0.04178  -5.561  < 0.001 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 
(Adjusted p values reported -- single-step method)


結論:可知F1(LA1543xLA1312)與LA1543形狀相似，但與LA1312和LA1589形狀有顯著不同。