Hi
I am kind of new of 16SrRNA data analysis.
I just was wondering if I have metadata with many variables , some of them would be expected to have a stronger influence than others on bacterial diversity and microbiome composition.
How I can adjust for those variables. For examples where I run functions like ---> alpha diversity ( estimate_richness()) or beta diversity (ANOSIM).
In order to adjust and distinguish which factors significantly influence( positively or negatively) the bacterial diversity,you need to run Redundancy analysis (RDA) or Principal component analysis(PCA).Through These analysis can be perform using softwares like Minitab,XLSTAT,R,or Canoco.Good luck
I don't know if I understand your question well. But as the answer before, if you want to explore the influence of your constrains on the microbiome structure the ANOSIM or Permanova will do what you're asking. Some considerations regarding microbiome data analysis, before doing these tests, keep in mind that the NGS data are compositional data, therefore you will need to perform some sort of transformation like clr, iqlr, mclr in order to properly analyze this kind of dataset. On the oder hand, if what you're looking for is to correlate your microbiome data with your metadata, ALDEx2 package allows you to do it with generalized linear model (glm) applied to compositional data.
Hi
I believe you are having Illumina Mi-seq data, if not, still it's fine.
1. Alpha n Beta diversity, both show the difference of treatment over the microbial community i.e. within a group or among the groups, respectively. Beta diversity can be analyzed in many ways such as NMDS or PCA, your choice.
2. Get the treatment-induced microbial abundance i.e. which microbe is majorly affected by treatment. you can use Kruskal-Wallis test or Lefse to get the abundance.
3. Run a correlation between the abundant microbes and metadata. OR, try to build a scientific logic behind these abundant microbes and variation in metadata. 4. A lot can be done, but for a real beginner, it's more than sufficient to start.
Cheers !!
Some suggestions may be helpful:
1. First, it is good to know if these variables were statistically different between your study groups.
2. Performing RDA/CCA to access the impact of those variables (environmental factors) on the microbial community composition.
3. This paper [doi:10.1136/gutjnl-2017-314281] is an example of dealing with confounding factors in 16S rRNA gene analysis.
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics