Table 1 The effect of statistical adjustment on performance.
From: DA DA: Degree-Aware Algorithms for Network-Based Disease Gene Prioritization
All Genes | Degrees ≤ m | Degrees >m | ||||
|---|---|---|---|---|---|---|
Method | Avg. Rank | AUROC | Avg. Rank | AUROC | Avg. Rank | AUROC |
Network Propagation | 22.75 | 0.77 | 30.00 | 0.70 | 15.57 | 0.84 |
Random walk w/restarts | 23.75 | 0.76 | 34.76 | 0.65 | 12.82 | 0.88 |
Based on seed degree | 21.42 | 0.78 | 21.51 | 0.79 | 21.35 | 0.79 |
Based on candidate degree | 19.50 | 0.81 | 21.20 | 0.79 | 17.82 | 0.82 |
Based on centrality | 20.40 | 0.80 | 24.35 | 0.75 | 16.47 | 0.84 |
- The effect of statistical adjustment on performance. Average Rank of the true disease genes and AUC values are listed. To demonstrate the effect of connectivity, we also provide separate results for the cases in which the weighted degree of true disease gene is ≤ m and > m where
is the average weighted degree of all genes (m = 4.23 in our experiments).
is the average weighted degree of all genes (m = 4.23 in our experiments).