About PCoA and principle component regression (PCR)

[angelrica_kuan]

Hi ~ I have few questions about PCA analysis after result was obtained from QIIME2.

There are many variable number for PCA (PC1, PC2, PC3....). Then, I would like to analyse the relationship between only "PC1" and specific " OTU/ASV" percentage.

(Huang, et al., 2016)

Is it possible to get the PC1 value from QIIME2 instead of using R? Which file I can find to obtain this value?

Thanks for your helping again.

Best wishes,
Angelrica Lee

[jwdebelius (Justine Debelius)]

Hi @angelrica_kuan,

The q2-diversity plugin runs PCoA. This is distance agnostic and starts from the distance matrix rather than the feature table; you lose the direct ASV/OTU association in processing. DECOIDE provides a robust PCA method; I would recommend going through their tutorial.

The factor loadings themselves can be exported from the PCA artifact.

Best,
Justine

PS I moved this to User Support because it better fits that topic

[angelrica_kuan]

Hi ~~~ Thanks for your helping !!