3 Analysis
3.1 WGCNA
In the demo1, we used the gene expression table in different samples and the trait data of samples.
In the analysis parameters:
- Specify Bicor to calculate the correlation by selecting the option Bicor
in the Aorrelation algorithms
parameter;
- Ahoose a type of network by selecting the option Signed
in the Network type
parameter;
- Specify 30 to check by inputting 30
in the textarea of Specify maximum power to check
parameter;
- Pick the top 2000 genes with the biggest difference between samples to analyse by inputting 2000
in the textarea of Only select top genes with maximum variances
parameter;
- Remove outlier samples by selecting TRUE
in the textarea of Remove outlier samples
parameter;
- Specify -2.5 as the threshold for judging abnormal samples by inputting -2.5
in the textarea of Outlier sample detection threshold
parameter;
- Specify 0.85 to define scale-free network by inputting 2
in the textarea of R square cut
parameter;
- Detect the module whose size is lower than 25 by inputting 25
in the textarea of Minimum module size for module detection
parameter;
- Specify 2 to split module by inputting 2
in the textarea of Deep split
parameter;
- Set a random number to ensure that the result is repeatable by inputting 2020
in the Seed for random number generator
parameter.
3.2 DE analysis (limma)
In the demo1, we used the raw reads count table in different samples and the metadata of samples.
In the analysis parameters:
Specify the type of input data by selecting the option
Raw reads count
in theExpression data type
parameter;Specify the column containing group information by selecting the option
conditions
in theGroup
parameter;Remove outlier samples by selecting
TRUE
in the textarea ofDetect and remove outlier samples
parameter;Do not detect and remove batch effect by selecting
False
in the textarea ofDetect and remove outlier samples
parameter;The threshold is 1 and -1 of Log2 Fold change by inputting
1
in the textarea ofLog2 Fold change
parameter;The threshold is 0.05 of FDR by inputting
0.05
in the textarea ofFalse discovery rate
parameter.