Table 3.
Predictor weights of the PLS model explaining the severity of acute diarrhea in children aged below five years in Kunming, southwest China
| Predictor variables | Loads | Weights | Percent | Cross-correlation |
|---|---|---|---|---|
| Virus richness | 0.68 | 0.66 | 43.2 | 0.95 |
| RVA | 0.54 | 0.56 | 31.4 | 0.61 |
| NoV GII | 0.45 | 0.43 | 18.8 | 0.42 |
| Ad | 0.14 | 0.14 | 2.0 | 0.04 |
| B. hominis | 0.05 | 0.09 | 0.8 | 0.0 |
| Protozoa richness | 0.05 | 0.09 | 0.8 | 0.0 |
| Bacteria richness | 0.09 | 0.08 | 0.7 | 0.02 |
| DEC | 0.07 | 0.07 | 0.5 | 0.01 |
| NTS | 0.06 | 0.04 | 0.2 | 0.01 |
| As | 0.11 | 0.02 | 0.0 | 0.02 |
| Cryptosporidium spp. | 0.0 | 0.01 | 0.0 | 0.0 |
| SaV | 0.05 | -0.01 | 0.0 | 0.0 |
Note: Predictor weights represent the contribution of pathogen to the PLS X’s component. Pathogens explaining more than 10 % of the observed PLS X’s block variability are shown in bold type. Cross-correlations represent the correlations between each pathogen and the MNLS per 24 h (diarrhea severity). Virus, protozoa, and bacteria richness signify the maximum number of species from each group detected in a child. Seasons have been defined as two periods: wet (May to October) and dry (November to April)