Skip to main content
. 2019 Jun 25;13(10):2589–2602. doi: 10.1038/s41396-019-0411-2

Table 1.

Top 10 variable loadings on the first three principal components of the PCA performed on the microbial traits dataset, shown in Fig. 1 and S2

PC1 Weight PC2 Weight PC3 Weight
ecosystemcategory_human −0.16 temperaturerange_mesophilic 0.19 growth_in_groups −0.24
specificecosystem_sediment 0.16 temperaturerange_thermophilic −0.19 gram_stain_positive −0.24
ecosystem_environmental 0.16 oxygenreq_strictanaero −0.19 cellarrangement_singles 0.21
knownhabitats_host −0.15 temperaturerange_hyperthermophilic −0.18 cellarrangement_filaments −0.20
ecosystemsubtype_intertidalzone 0.15 knownhabitats_hotspring −0.17 sporulation −0.20
ecosystem_hostassociated −0.15 exosystemtype_rhizoplane 0.17 energysource_chemoorganotroph −0.19
habitat_hostassociated −0.15 habitat_specialized −0.16 cellarrangement_clusters −0.18
habitat_freeliving 0.15 metabolism_methanogen −0.16 shape_tailed −0.18
ecosystemtype_digestivesystem −0.14 ecosystemcategory_plants 0.15 habitat_terrestrial −0.18
specificecosystem_fecal 0.14 ecosystemtype_thermalsprings −0.15 motility 0.17

These three components explain 17%, 10%, and 7% of the total variance, respectively