next up previous contents
Next: 3.3.11 Correlation Analysis Up: 3.3 Data analysis Previous: 3.3.9 Signal Histogram   Contents

3.3.10 Differential Analysis

The aim of the differential analysis is to compare the signal value of each probe between two user-defined groups of tumors. The user has to perform the following steps:

Figure 3.61: Views $\rightarrow $ New View $\rightarrow $ Double View $\rightarrow $ Point View - First the user must open a new double window.
Image DIFFANA0

Figure 3.62: File $\rightarrow $ Import $\rightarrow $ in Top Panel - Then the user must loading several profiles in the top panel.
Image DIFFANA1

Figure 3.63: File $\rightarrow $ Import $\rightarrow $ in Bottom Panel - Then the user must loading several profiles in the bottom panel.
Image DIFFANA2

Figure 3.64: Tools $\rightarrow $ Top Panel $\rightarrow $ Differential Analysis $\rightarrow $ Compute - The user can open a new window of dialog for Differential Analysis.
Image DIFFANA3

Different options are available within the Differential Analysis dialog box:

Test:
Statistical test is performed using either Student or Welch test.
Color bars Use sign:
If the probe mean of the group in the bottom panel is greater than the probe mean of the group in the top panel, the bar is displayed in red and otherwise in green
Color bars Use confidence:
Color bar corresponds to the percentage of profiles without missing values
Max P-Value:
Probes with a p-value greater than this value are discarded
Multiple testing Benjamini-Hochberg:
ajusted p-values using Benjamini-Hochberg method
Multiple testing Benjamini-Yekutieli:
ajusted p-values using Benjamini-Yekutieli method
Results:
The results are displayed in a profile view (see Figure 3.66). The values correspond to $-\log_10$p-value
Report:
The results are saved in HTML or CSV file

Figure 3.65: Differential Analysis dialog - Different options are available.
Image DIFFANA4

Figure 3.66: Differential Analysis profiles - Differential Analysis view.
Image DIFFANA5


next up previous contents
Next: 3.3.11 Correlation Analysis Up: 3.3 Data analysis Previous: 3.3.9 Signal Histogram   Contents
2007 - Institut Curie Bioinformatics unit