3.2.1.10 TCGA-PAAD dataset

There are 36886 genes and 182 samples in the dataset.

At least some samples or genes look not suitable for the co-expression analysis. After removing them, there remain 36870 genes and 182 samples.

##    Power SFT.R.sq slope truncated.R.sq  mean.k. median.k. max.k.
## 1      1    0.278 -1.74          0.888 4.36e+03  4.30e+03 8270.0
## 2      3    0.724 -2.06          0.968 2.58e+02  1.92e+02 1340.0
## 3      5    0.810 -2.17          0.988 3.88e+01  1.73e+01  427.0
## 4      7    0.881 -2.05          0.997 9.62e+00  2.27e+00  193.0
## 5      9    0.905 -1.97          0.998 3.27e+00  3.84e-01  115.0
## 6     11    0.926 -1.84          0.999 1.39e+00  7.93e-02   76.4
## 7     13    0.932 -1.73          0.996 6.94e-01  1.87e-02   54.0
## 8     15    0.926 -1.64          0.989 3.90e-01  4.86e-03   39.8
## 9     17    0.919 -1.57          0.981 2.39e-01  1.35e-03   30.3
## 10    19    0.945 -1.48          0.995 1.57e-01  3.95e-04   23.6
## 11    21    0.956 -1.41          0.997 1.09e-01  1.20e-04   18.7
## 12    23    0.964 -1.35          0.990 7.83e-02  3.81e-05   15.0
## 13    25    0.983 -1.29          0.993 5.83e-02  1.22e-05   12.1
## 14    27    0.979 -1.29          0.991 4.47e-02  4.02e-06   11.0
## 15    29    0.971 -1.29          0.987 3.51e-02  1.35e-06   10.3

Network is computed using power=7.

Here are the modules for a few genes of interest:

• VNN1 is in the cyan module.
• PPARA is in the grey module.
• PPARD is in the grey module.
• PPARG is in the royalblue module.
• PPARGC1A is in the cyan module.
• PPARGC1B is in the yellow module.

g:Profiler results for all modules are available here. Since there are many modules, you should expect at least 1 minute computation on the g:Profiler website for this query. Check below for faster queries focused on VNN1 module.

Figure 3.75: Pathway

Figure 3.76: Kegg

Figure 3.77: MSigDb Hallmarks. readable pdf version.

## No enriched DO term in these classes

To check that module-to-genes reattribution is working as expected, heatmap of signal for VNN1 module is produced, both as absolute values and as Z-score ones.

Figure 3.78: log2(vst) signal in VNN1 module. Rows were gene name is prefixed with "-" display negative Z-score for these genes, meaning their signal is anticorrelated with other genes in the cluster. pdf version with readable genes and findable VNN1.

Figure 3.79: Zscore of log2(vst) signal in VNN1 module. Rows were gene name is prefixed with "-" display negative Z-score for these genes, meaning their signal is anticorrelated with other genes in the cluster. pdf version with readable genes and findable VNN1.

Figure 3.80: Genes in VNN1 module plotted by Pearson correlation against it. Details are available in this table.

• Enrichr results for all genes in VNN1 module are available here
• Enrichr results for genes in VNN1 module and positively correlated with it are available here
• Enrichr results for genes in VNN1 module and negatively correlated with it are available here
• g:Profiler results for all genes in VNN1 module are available here
• g:Profiler results for genes in VNN1 module and positively correlated with it are available here
• g:Profiler results for genes in VNN1 module and negatively correlated with it are available here
• g:Profiler results for comparison of the three queries above are available here