search

mortonjt / ancomP

Analysis of Composition of Microbiomes
11 stars 2 forks source link

continuous metadata and adjusting covariates #4

Open o147942 opened 7 years ago

o147942 commented 7 years ago

Hi, I found ANCOM from QIIME2. I'd like to use ANCOM to identify associations between taxa and continuous sample metadata (such as PH). I used the MaAsLin before I know the ANCOM. Is the linear regression analysis possible in ANCOM? I need to adjust covariates such as sex and age. I think the ANCOM is very good tool, so I want to use it for my study. In the paper of ANCOM, I read that the ANCOM can be implemented in a linear model framework to adjust for covariates. I'm not a expert in both statistics and developing code.. T.T Manual of ancom.R is also difficult for me.. How can I implement a linear regression as well as using covariates in ANCOM?

Thank you. Hanna

mortonjt commented 7 years ago

Thank you for your interest @o147942 .

Its been a while since I looked at the internals of ANCOM - I do not know if it can handle linear regression. Personally, I have been moving away from using ANCOM due its strict assumptions.

However, I have developed another technique that has fewer assumptions and can definitely handle linear regression. A tutorial on how to do this can be found here. If you have any more questions, let me know!

o147942 commented 7 years ago

Thank you for your kindly reply @mortonjt . I'll try your new suggestion.