Question: Differential taxonomic analysis using longitudinal model and siamcat

zellerlab / siamcat

R package for Statistical Inference of Associations between Microbial Communities And host phenoType

https://siamcat.embl.de/

50 stars 16 forks source link

Question: Differential taxonomic analysis using longitudinal model and siamcat #45

Open alienzj opened 5 months ago

alienzj commented 5 months ago

Dear Siamcat team,

may I know how to do differential taxonomic analysis using longitudinal model and siamcat?

For example, data show below:	id	group
a	normal	T1
a	normal	T2
b	disease	T1
b	disease	T2

How to compare the difference of taxonomic composition of normal and disease group across T1 and T2 ? And we know for subject a or b, they may have different taxonomic composition at T1 and T2 regards of normal and disease condition.

jakob-wirbel commented 5 months ago

Hi @alienzj Thank you for using SIAMCAT! I think there are several things that you could do:

you could look at the differences between T1 and T2. For each species and patient, you could calculate the change that happened between time points and then use this table of differences as input for SIAMCAT and then calculate associations between groups
alternatively, you could use a random effects model within SIAMACT. For example, you could use id as a random effect like this: check.associations(sc.obj, formula='feat~label+(1|id)'). This will fit a random effect model to each feature and output the significance for the difference in label (group in your example), with a random effect for the patient ID

Let me know if that works! :)

alienzj commented 5 months ago

Dear @jakob-wirbel ,

Thank you so much for your suggestions. Let me try it.