Hey all,
we are doing microbiome analysis with soft tissue of crayfish. We perform all our experiments in biological triplicates, means that we have three crayfish per timepoint.
We extract the DNA from the soft tissue and send the extracted DNA to a sequencing facility that sequence each of the replicates individually. We get the raw data, analyse them via QIIME2 and after the taxonomic classification, I proceed working with R.
So, my question is, how would you work with these replicates, analyse the replicates, respectively?
Do you take the mean of the counts, e.g. when you analyse the microbial community in the samples? In that case, do you also look at the stdev? And how? Sometimes the replicates are very different. Do you have any recommendation on programs that analyse the stdev of the replicates? Or tell me if the replicates are comparable?
Or do you take each replicate alone? Another possibility would be to take only one of the replicates for the analysis? But how do you decide which one?
You would help me a lot,
Kathrin
use one way analysis of variance time point x replicate to compare variance beween time point with variance between replicates.
take each replicate alone Another possibility would be to take only one of the replicates for the analysis.
Dear Kathrin,
I guess that your final goal is to obtain a list of differential abundant OTUs between time-points, is that right? There are some tools that performs this kind of analysis, such as ANCOM, but due to the low number of replicates for each time-point it wouldn't be feasible. ANCOM has a low sensitivity with a low number or replicates. A good alternative would be the use of DESeq2 in order to perform pair-wise comparisons between time-points. Although DESeq2 was firstly developed for the analysis of RNA-seq data, several benchmarks have tested it on microbiome data with good results.
In case your R skills are poor, you could do a microbiome analysis in an online tool called Gaia (https://metagenomics.sequentiabiotech.com/), which is user-friendly, infers the taxonomic composition of the samples and performs differential analyses with DESeq2.
Best regards,
Andreu Paytuví-Gallart
Bioinformatician @ www.sequentiabiotech.com
- Badges
- Science topic
- Similar topics
- Engineering
- Chemical Engineering