how to find ASVs present in different "sites"

[callmcgovern]

Hello, I am trying to look at the ASVs that are similar between different sites. I have sites in my sample_data and I have been able to make a venn diagram of overlapping vs non-overlapping ASVs in different sites. However, I would now like to look at a list of those ASVs which are shared by two particular samples. Does anyone know how I could do this? Thanks!

[jeremysutherland]

> ps.rare
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 2864 taxa and 893 samples ]
sample_data() Sample Data:       [ 893 samples by 10 sample variables ]
tax_table()   Taxonomy Table:    [ 2864 taxa by 7 taxonomic ranks ]
> 
> ps.NY <- subset_samples(ps.rare, site == "NY")
> ps.NY<- prune_taxa(taxa_sums(ps.NY) > 0, ps.NY) # prune OTUs that are not present in at least one sample
> 
> ps.NJ <- subset_samples(ps.rare, site == "NJ")
> ps.NJ <- prune_taxa(taxa_sums(ps.NJ) > 0, ps.NJ) # prune OTUs that are not present in at least one sample
> 
> ps.NY
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 951 taxa and 349 samples ]
sample_data() Sample Data:       [ 349 samples by 10 sample variables ]
tax_table()   Taxonomy Table:    [ 951 taxa by 7 taxonomic ranks ]
> ps.NJ
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 1141 taxa and 204 samples ]
sample_data() Sample Data:       [ 204 samples by 10 sample variables ]
tax_table()   Taxonomy Table:    [ 1141 taxa by 7 taxonomic ranks ]
> 
> length(taxa_names(ps.NY))
[1] 951
> length(taxa_names(ps.NJ))
[1] 1141
> length(intersect(taxa_names(ps.NY), taxa_names(ps.NJ)))
[1] 317
>write.csv(intersect(taxa_names(ps.NY), taxa_names(ps.NJ)), "shared_asvs_NY_NJ.csv")

[callmcgovern]

Hi, thank you for your help! How would you do this if you wanted to find ASVs that are found in all of your samples in the phyloseq object?

[jeremysutherland]

nrow(ps@otu_table) ps.prune <- prune_taxa(taxa_sums(ps) > 0, ps) # prune ASVs that are not present in at least one sample nrow(ps.prune@otu_table)

[callmcgovern]

Hi Jeremy, thanks for your quick reply. This does not seem to work. Here is the code and output on my end:

nrow(ps_dolphin@otu_table) [1] 15

ps_dolphin phyloseq-class experiment-level object otu_table() OTU Table: [ 1899 taxa and 15 samples ] sample_data() Sample Data: [ 15 samples by 4 sample variables ] tax_table() Taxonomy Table: [ 1899 taxa by 7 taxonomic ranks ] refseq() DNAStringSet: [ 1899 reference sequences ]

ps.prune <- prune_taxa(taxa_sums(ps_dolphin) > 0, ps_dolphin) nrow(ps.prune@otu_table) [1] 15

ps.prune phyloseq-class experiment-level object otu_table() OTU Table: [ 1899 taxa and 15 samples ] sample_data() Sample Data: [ 15 samples by 4 sample variables ] tax_table() Taxonomy Table: [ 1899 taxa by 7 taxonomic ranks ] refseq() DNAStringSet: [ 1899 reference sequences ]

Any ideas as to why it wouldnt be working?

[kodelogue]

Could be that you don't have any ASVs with all zero counts. (All ASVs have at least 1 count in at least one sample).

This should tell you if thats the case. (Assuming samples are rownames)

sum(colSums(ps_dolphin@otu_table)<1)

[jeremysutherland]

ope. Kodelogue is my other (fun) account. but that should do the trick.

[mailasb]

Hi, I'm new to Github. I found this answer that was very helpful for my project:

> ps.rare
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 2864 taxa and 893 samples ]
sample_data() Sample Data:       [ 893 samples by 10 sample variables ]
tax_table()   Taxonomy Table:    [ 2864 taxa by 7 taxonomic ranks ]
> 
> ps.NY <- subset_samples(ps.rare, site == "NY")
> ps.NY<- prune_taxa(taxa_sums(ps.NY) > 0, ps.NY) # prune OTUs that are not present in at least one sample
> 
> ps.NJ <- subset_samples(ps.rare, site == "NJ")
> ps.NJ <- prune_taxa(taxa_sums(ps.NJ) > 0, ps.NJ) # prune OTUs that are not present in at least one sample
> 
> ps.NY
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 951 taxa and 349 samples ]
sample_data() Sample Data:       [ 349 samples by 10 sample variables ]
tax_table()   Taxonomy Table:    [ 951 taxa by 7 taxonomic ranks ]
> ps.NJ
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 1141 taxa and 204 samples ]
sample_data() Sample Data:       [ 204 samples by 10 sample variables ]
tax_table()   Taxonomy Table:    [ 1141 taxa by 7 taxonomic ranks ]
> 
> length(taxa_names(ps.NY))
[1] 951
> length(taxa_names(ps.NJ))
[1] 1141
> length(intersect(taxa_names(ps.NY), taxa_names(ps.NJ)))
[1] 317
>write.csv(intersect(taxa_names(ps.NY), taxa_names(ps.NJ)), "shared_asvs_NY_NJ.csv")

I wanted to ask: How could I obtain that same list of shared ASVs between both groups but also associated with the taxonomic classification found in the tax_table? In other words, I want to obtain a table (like the one in the attached image) that shows each shared ASV (for example: ASV1, ASV5, ASV30, etc.) and what microorganism it represents. Thank you very much.

[jeremysutherland]

I think* this should work:

asv_list <- intersect(taxa_names(ps.NY), taxa_names(ps.NJ))

taxa_sub <- ps.rare@tax_table[rownames(ps.rare@tax_table) %in% asv_list, ]

[mailasb]

It worked perfectly. Thank you very much!