Table 5.
As per Tables 3 and 4, with genes-and-microbes model data. Top ranking features with p-values less than 0.05 and their importance scores discovered by our genes-and-microbes model (Left). Side-paired differential expression (fold change) analysis results of TPM and CPM values for the same features (Right) Wilcoxon-rank sum test was used to calculate p-values and FDR (Benjamini & Hochberg)
Model Feature Importance Metrics | Differential Expression | |||||||
---|---|---|---|---|---|---|---|---|
Rank | ENSG ID_Gene/Tax ID_Name | Importance Score | Log Importance Score | p-value | Log2 FC | p-value | FDR | Associated Side |
1 | ENSG00000142661_MYOM3 | 0.037 | -3.29 | 0 | -0.62 | 1.36E-08 | 2.71E-08 | Left |
2 | ENSG00000198353_HOXC4 | 0.033 | -3.40 | 0 | 1.88 | 2.11E-15 | 3.23E-14 | Right |
3 | 33043_Coprococcus eutactus | 0.027 | -3.62 | 0 | 2.08 | 2.00E-14 | 1.54E-13 | Right |
4 | ENSG00000159182_PRAC1 | 0.024 | -3.72 | 0 | -2.86 | 5.08E-21 | 2.34E-19 | Left |
5 | ENSG00000260597_AC012531.25 | 0.020 | -3.90 | 0 | 1.19 | 4.26E-12 | 2.17E-11 | Right |
6 | 33038_[Ruminococcus] gnavus | 0.016 | -4.13 | 0 | 2.07 | 1.31E-15 | 3.02E-14 | Right |
7 | ENSG00000184719_RNLS | 0.011 | -4.50 | 0.01 | -0.89 | 2.85E-10 | 9.37E-10 | Left |
8 | ENSG00000197757_HOXC6 | 0.011 | -4.54 | 0 | 1.28 | 5.02E-12 | 2.31E-11 | Right |
9 | ENSG00000144451_SPAG16 | 0.010 | -4.61 | 0 | -0.64 | 9.00E-09 | 1.97E-08 | Left |
10 | 851_Fusobacterium nucleatum | 0.010 | -4.62 | 0 | 1.67 | 2.73E-06 | 3.49E-06 | Right |
11 | ENSG00000273374_RP11-383I23.2 | 0.010 | -4.62 | 0 | -1.02 | 2.86E-08 | 5.47E-08 | Left |
12 | 446043_uncultured Lachnospira sp. | 0.010 | -4.63 | 0 | 1.46 | 5.97E-09 | 1.37E-08 | Right |
13 | 165179_Prevotella copri | 0.009 | -4.66 | 0 | 1.6 | 1.50E-09 | 3.85E-09 | Right |
14 | 154288_Turicibacter sanguinis | 0.009 | -4.67 | 0 | 1.9 | 5.00E-08 | 8.51E-08 | Right |
15 | 59620_uncultured Clostridium sp. | 0.009 | -4.67 | 0 | 1.09 | 6.86E-12 | 2.87E-11 | Right |