Table 5.
Effect of bin sizes in random sampling on pcent calculation. The table shows the Pearson correlations between the pcent values calculated using pairs of bin sizes (b) from the set {10,20,40}. The pcent values are highly correlated between different bin sizes (Python Scipy p-values for the correlations are below 0.0001 for all pairs).
Pearson Correlation | ||||
---|---|---|---|---|
Asthma | BPD | COPD | ||
GO BP | pcent(b = 10) vs. pcent(b = 20) | 0.9840 | 0.9664 | 0.9652 |
pcent(b = 20) vs. pcent(b = 40) | 0.9903 | 0.9848 | 0.9620 | |
KEGG | pcent(b = 10) vs. pcent(b = 20) | 0.9902 | 0.9982 | 0.9806 |
pcent(b = 20) vs. pcent(b = 40) | 0.9793 | 0.9762 | 0.9982 |