
Df

Sum Sq

Mean Sq

F value

Pr(>F)


Peptide

12

170916

14243.0

57461.65

< 2.2e16

Subarray

2

24

12.2

49.18

< 2.2e16

Needle

15

45

3.0

12.08

< 2.2e16

Row

234

458

2.0

7.89

< 2.2e16

Column

75

409

5.5

22.03

< 2.2e16

Residuals

2218

550

0.2
  
 Linear model fit summary. The Fvalues and the corresponding probabilities (Pr(>F)) clearly indicate that all explanatory variables used are highly significant and that each contributes to reducing the variance present in the arrays. The peptide sequence itself shows by far the strongest effect while requiring only twelve degrees of freedom (Df), it shows a sum of squares for its effect (Sum Sq) of 170,916, much stronger than the residuals sum of squares of 550. While not as strong as the peptide effect, the remaining explanatory effects reduce the residuals by 65%. Column and row effects are strongest, but subarray and needle effects require fewer degrees of freedom, resulting in strong mean sum of squares (Mean Sq) values.