Re: questions about MaAsLin analysis using Galaxy

[Ayshwarya Subramanian]

Hi Maria, 

1) The default Maaslin run in Galaxy expects values between 0 and 1. When using that, either CSS or dividing by sum of values is fine, not both. 

2) It seems Maaslin is more appropriate based on your email. Here's some more information to think about your analyses:

LEfSe identifies those data features that are distinct between a pair of metadatums (e.g. differences between two sampling sites, two clinical outcomes, two biochemical markers, two modalities, etc.).  MaAsLin extends the functionality of LEfSe to identify associations between data features and multiple metadata factors, which can be discrete and/or continuous and can include time series data.

On Fri, Oct 9, 2015 at 1:55 PM, Maria Glymenaki <maria.g...@manchester.ac.uk> wrote:

Hello,

I am a new user to MaAsLin and I have a couple of questions. I want to ask you about the normalization process you mention in the Galaxy website. I looked on the forum website and another user had performed CSS normalization on the data coming from OTU table and then each value on each sample was divided by the sum of all the values in that sample. I want to ask you if this is normalization approach is appropriate for this kind of analysis. The sequencing depth is different in different sample, I need somehow to make the results comparable. Another way is applying rarefaction, but it will be at the expense of low abundance OTUs.

Second, I want to ask you about the difference between MaAsLin and LefSE. I read the initial publications, but I got confused in what type of analysis applies to my data. I have microbiome data derived from rodents with the following metadata : two time points ( A and B etc), two different locations (C and D etc), two genotypes, different response to one particular treatment (of the type responder or non-responder), different litters and different cages . I thought of doing MaAsLin to identify correlation of microbes with metadata categories. If I want to see if specific microbiota characterize a specific condition ( time point A and location C), shall I use LefSE or shall I modify the initial input file I use for MaAsLin ? Any clarification will be much appreciated.

I would like to thank you for your time and assistance.

Regards,

Maria Glymenaki

Maria Glymenaki BSc, MSc

PhD student in Immunology

AV Hill Building

Faculty of Life Sciences

University of Manchester

Oxford Road,  M13 9PT, UK

phone: + 161 30 64221

email: maria.g...@manchester.ac.uk

 

[Han-na Kim]

Hi,

If I want to use CSS data in Maaslin, how can I use it?

As you said, the default Maaslin use values between 0 and 1.

I use Maaslin in R.

Let me know the option for using CSS data, which is not value between 0 and 1.

Thank you.

Hanna

2015년 10월 16일 금요일 오전 12시 58분 13초 UTC+9, subraman 님의 말:

[Han-na Kim]

One more question!

I don't know why I cannot use both CSS and dividing by sum (the first CSS normalization then dividing by sum in CSS normalized data).

Can you explain it?

I think that just dividing by sum is not enough to nomalize my sequence data.

Regards,

Hanna

2016년 11월 11일 금요일 오후 7시 32분 10초 UTC+9, Han-na Kim 님의 말: