TimeOmics on repeated cross-sectional data

Analysis
1

Hi there,

I read about TimeOmics and it seems to be possible to use it on my poultry dataset. In the dataset, I have two treatment + one control group and four time points (age day 3, day 14, day 21 and day 35). For each time point, in each condition, there are 4 biological replicates. For each replicate, we measured the paired gut metagenomics and poultry mucosa RNA-seq. I would like to identify treatment-specific associations between microbes/microbial gene clusters and host transcripts across the time.

I know TimeOmics is developed for longitudinal data but my dataset is not longitudinal. At each time point, those chicken were sacrificed to get the paired data so this dataset is actually repeated cross-sectional. So I would like to discuss this with you if TimeOmics can still be used in this case. If so, how can I use it correctly, e.g., in the time modelling step, should I try model each feature only with linear model and LMMS with defined basis?

I looked through all the posts about TimeOmics but didn’t find anyone asking about repeated cross-sectional data. I’m very looking forward to hear your insights on this.

Best regards,
Burrita

2

hi @buritta,

The only way to have timeOmics applied would be to match the individuals across time (i.e similar characteristics). To be a bit more confident in your analysis, you could also swap those individuals several times when you assign them as a ‘unique individual ID’ and check that the results are similar in the profiles you obtain etc. If you explain this clearly in your analysis, and show the results are similar, then you can justify making those approximations.

Kim-Anh

1 Like