Keeps the first copy of every distinct sequence, like `seqkit rmdup -s`.
Details
Be careful when de-replicating a *reference* ("good") database that is used by [clean_reference()]: if two genera share an identical sequence, keeping only one copy hides exactly the ambiguity that the "matches multiple genus" check is designed to detect. De-replicating the contaminant database is harmless.
Examples
seqs <- c(a = "ACGTT", b = "ACGTT", c = "AACGT", d = "GGGCC")
dereplicate(seqs)
#> a d
#> "ACGTT" "GGGCC"
dereplicate(seqs, both_strands = FALSE)
#> a c d
#> "ACGTT" "AACGT" "GGGCC"