Multiple markers/loci per cluster issue

[bpjburger]

Hello,

I ran PhylotaR according to the tutorial, modifying it as is shown in the code below. To create fasta files for each cluster I ran a loop in R. But when checking the sequences in the resulting fasta-files I discovered that they contain different loci. The first cluster (and corresponding fasta) contained ITS and trnL. This issue is present in other clusters as well even in the pre-filtered phylota-object.

Is this the result of a mistake in my code or is this a PhylotaR issue? And how can I possibly fix it?

Regards,

Bart

The code:

library(phylotaR)
wd <- '[FILEPATH TO WORKING DIRECTORY]'
ncbi_dr <- '[FILEPATH TO COMPILED BLAST+ TOOLS]'
txid <- 4210  # Asteracea ID
setup(wd = wd, txid = txid, ncbi_dr = ncbi_dr)
run(wd=wd)
phylota=read_phylota(wd)

#Create a new Phylota-object that has one sequence per taxon.

phylota_sp <- drop_by_rank(phylota = phylota, rnk = 'species', n = 1)
cids=phylota_sp@cids

#Create a new Phylota-object, but with excluding clusters that only contain a few taxa. 

ntaxa=get_ntaxa(phylota = phylota_sp, cid = cids)
keep=cids[ntaxa>50]#new set of cluster IDs with more than 50 taxa
selected=drop_clstrs(phylota = phylota_sp, cid = keep)

#And now the loop.

cids_sel=selected@cids

#A for loop that goes over the cluster IDs
for (i in cids_sel){
  txids <- get_txids(phylota = selected, cid = i, rnk = 'species')#makes a list of taxon-IDs for the specific cluster

# look up name for Taxon IDs per cluster and create a list out of it.
scientific_names <- get_tx_slot(phylota = selected, txid = txids, slt_nm = 'scnm')

#formatting the names
scientific_names <- gsub('\\.', '', scientific_names)
scientific_names <- gsub('\\s+', '_', scientific_names)

#specify the sequence IDs per cluster
  sids <- reduced@clstrs[[i]]@sids

  # write out (one file per cluster)
  pad=[GENERAL PATH TO LOCATION]
  outfile= str_c(pad, i)#add the cluster ID to the filename
  write_sqs(phylota = selected, sid = sids, sq_nm = scientific_names,
            outfile = outfile)}

[DomBennett]

Hi,

I've managed to re-run your analysis for Asteracea and I'm looking over the clusters. So that I can repeat your process, how are you determining that there is a mix of ITS and trnL in the first cluster?

The top thing I've noticed is ...

#specify the sequence IDs per cluster
sids <- reduced@clstrs[[i]]@sids

Should be:

#specify the sequence IDs per cluster
sids <- selected@clstrs[[i]]@sids

Where does the "reduced" phylota object originate? Mixing sids across different phylota objects could be causing the mixing of sequences in a cluster.

[Bunholi]

Hi @bpjburger did you solve the problem following @DomBennett correction? I am planning to do the same around here

[bpjburger]

Hi @Bunholi,

Unfortunately, the corrections did no solve the problem. I think that it is a software issue.

[maelle]

For info, we're looking for a new maintainer / a new maintainer team for this package, see #57 and feel free to volunteer, we'd be happy to help.