Cluster ASVs to OTUs with DECIPHER

[Danyang1111]

Hi,
I am trying to cluster obtained ASVs to OTUs using DECIPHER, the code is as follows:
`
asv_sequences <- colnames(seqtab_nochim_subset)
sample_names <- rownames(seqtab_nochim_subset)

dna2 <- Biostrings::DNAStringSet(asv_sequences)

aln <- DECIPHER::AlignSeqs(dna2, processors = nproc)
d <- DECIPHER::DistanceMatrix(aln, processors = nproc)

clusters <- DECIPHER::TreeLine(
myDistMatrix=d,
method = "complete",
cutoff = 0.03, # use cutoff = 0.03 for a 97% OTU
processors = nproc)

merged_seqtab <- seqtab_nochim_subset %>%
t %>%
rowsum(clusters$cluster) %>%
t
`
and I got error like this:

Error in h(simpleError(msg, call)) : error in evaluating the argument 'x' in selecting a method for function 't': incorrect length for 'group'

what should I do?

[benjjneb]

In order to diagnose what is going wrong, I would start by breaking your multi-call line into individual steps to identify where the error is coming from.

val1 <- seqtab_nochim_subset %>% t
val2 <- rowsum(val1, clusters$cluster)
val3 <- t(val2)

After identifying which individual step is causing the error, you can inspect the help for the individual function (rowsum or t) and the input to the function to see if there are any mismatches that would cause the error.

[sigrik92]

Hi, I encountered the same issue and solved it by adding type = "clusters" (else the output is a dendrogram):

clusters <- DECIPHER::TreeLine(
myDistMatrix=d,
method = "complete",
type = "clusters",
cutoff = 0.03
processors = nproc)