A Practical Guide to Cancer Systems Biology

60 A Practical Guide to Cancer Systems Biology

3. Set “Transcripts” to the “merged transcripts” generated by Cuffmerge.


4. In “Condition” regions, set “Name” to the proper label (e.g., MiaPaCa2)
   and set “Replicates” input to the read-aligned SMA/BAM file generated
   by TopHat or HISAT for each of the RNA-seq conditions. If your
   experiment has more than two conditions, include additional conditions
   by clicking “Insert Condition”.


5. If using only one sample per condition, “Dispersion estimation method”
   must be “blind”.


6. Set “False Discovery Rate” to “0.05” or less.


7. Click “Execute” to run the analysis.

Visualize Cuffdiff output via CummeRbund

CummeRbund can visualize and integrate all of the data generated by
a Cuffdiff analysis which helps the users to explore their expression
data.

1. In the Tools panel, expand “NGS: RNA Analysis” and click “cummeR-
   bund”.


2. Set “Select backend database (sqlite)” input to the SQLite file generated
   by Cuffdiff.


3. Click “Insert Plots” and set “Plot type” of your interest.


4. Click “Execute” to generate the plot.

Select differentially expressed genes

Finally, we can use Filter function provided by Galaxy to selected differen-
tially expressed genes.

1. In the Tools panel, expand “Filter and Sort” and click “Filter”.


2. Set “Filter” to the file “gene differential expression testing” generated by
   Cuffdiff.


3. Set “With following condition” to “c14 == ‘yes’ ” where c14 denotes the
   14 thcolumns of the table.


4. Set “Number of header lines to skip” to “0”.


5. Click “Execute” button.


6. A file will be generated in history panel and click the eye icon to view
   the dataset. In addition, this file can be used in further analysis, such as
   functional or pathway analyses.