Metadata Preparation for MaAsLin2 Input

Downstream analysis and statistics MaAsLin
1

Dear Sir,

My name is Aditi Jangid, research scholar from India. I am working on metagenomic data. My aim is to explore the change in antibody expression with fecal microbiome community (longitudinal study). For the same I have prepared the metadata for MaAsLin2 and put antibody expression data on the metadata sheet as it is viz, absolute value (although my metagenomic data is normalized). Now, I need to know whether this is the correct method or do I need to normalize the antibody data first (similar way of normalization as microbiome data). For your reference find the metadata sheet attached. I request you to please provide your valuable suggestions. Your help is highly appreciated.

Thanks & regards

Aditi Jangid
Research Scholar
BioX Centre, School of Basic Sciences
Indian Institute of Technology, Mandi, India

2

Dear Dr. Jangid,
Thanks for reaching out. A main concern for continuous valued metadata such as antibody expression in MaAsLin2 is outliers. In your provided metadata it seems that the expression values have fairly large dynamic range (~15 fold change from min to max), but values are reasonably evenly distributed in this range, with few outliers. I’d say using these values unnormalized should be okay. Another option is to log transform them first. You can probably try both and see if either yields more meaningful results!
Thanks,
Siyuan