tax_table questions

Other Bioinformatics Tools
#1

I have imported data from qiime2.

And this is the code that I used to import qiime to phyloseq
ps.ng.tax <- read_phyloseq(otu.file = “table_taxonomyV2.biom”,
taxonomy.file = NULL,
metadata.file = “metadata-renew.csv”,
type = “biom”)

I have question. When I look up the tax table in species level,

table(tax_table(log.ps)[,“Species”])
image
For example I have 4 Campanella. I thought this means 4 differenct ASVs are annotated with one Campanella. right?
However my coworker told me that only 1 has to be written instead of 4. He is saying identical species have to be collapsed.

So the question is When I import feature table(=otu table) from qiime2, should I collapse and then import to phyloseq?
Or is it okay to use table from DADA2?

Can you tell me the suggestion?

1 Like
#2

Good afternoon!

That’s right!

You can collapse ASVs at the species level, turning your ASV table into a species table… but you don’t have to! There are pros and cons to both methods. :thinking:

Yes! Keeping your ASVs separate could let you resolve within-species variation of the your microbes (1, 2).

But if you would like to merge at the species level, you can do that using tax_glom().

log.ps.species <- tax_glom(log.ps, taxrank="Species")

Let me know what you decide to do!
Colin

#3

Thanks alot colinbrislawn!

In my lab, I’ve debated this issues with my coworker.
Main topic is that ASVs represent exact species. So as you mention it, even if several ASVs represents one species, it could be withing-species variation. However, in the microbiome like bacteria, 1% difference means they are different species. But in my case, I am analyzing zooplankton COI seqs. And several paper says up to 3% seqs differences are allowed. So in my lab, we regard those species as a identical species(Usually our baseline is 1.2%).

So if I put on the strict standard, I might use ASVs sparated so that I could do alpha and beta diversity more precise citeria. On the other hand, if I collapse ASVs at the species level, can I say I might analyze with more flexible standard?

Both way p-values were below 0.05

1 Like
#4

I have never studied zooplankton. Can you tell me more about how the species concept is applied to these microbes?

Do you think the COI sequence would show more or less variance across taxa and within the same taxa as the V4 16S sequence used in the Qiime2 tutorials?

Colin