Home Software Services About Contact usearch manual
What's new in USEARCH v9.2, 9.1 and 9.0

9.2 features
Calculate alpha and beta diversity metrics.
  New OTU table commands.

9.2 commands
alpha_div  Calculate alpha diversity metrics
  beta_div  Calculate beta diversity metrics, clusters and trees
  otutab_norm  Normalize OTU table to same number of reads per sample
  otutab_counts2freqs  Convert OTU table from counts to frequencies
  otutab_freqs2counts  Convert OTU table from frequencies to counts
  otutab_trim  Trim low-abundance counts, OTUs and samples from an OTU table
  otutab_otu_subset  Extract a subset of OTUs
  otutab_sample_subset  Extract a subset of samples
  otutab_stats  Report summary statistics for an OTU table
9.1 algorithm
UNCROSS detection and filtering of cross-talk.
  Roughly 2% of reads are assigned to the wrong sample.
  The UNCROSS algorithm can detect and filter cross-talk.
  See UNCROSS paper for details.

9.1 commands
  uncross  Detect and filter cross-talk in a OTU table.
  unoise2  Updated version of unoise using the UNOISE2 algorithm (Edgar 2016).

9.0 algorithms
UNOISE error-correction (denoising) paper:
SINTAX taxonomy prediction  paper:
  UCHIME2 chimera detection  paper:

9.0 commands
  annot  Annotate sequences from marker gene experiment (mock or real).
  calc_distmx_smallmem  Reduced-memory version of calc_distmx.
  fastx_demux  Demultiplex reads by matching index reads to barcodes.
  fastx_info  Summarize contents of FASTA/Q file.
  fastx_learn  Determine error rates of amplicon reads.
  fastx_uniques  Find unique sequences and abundance (new name for derep_fulllength).
  filter_phix  Filter reads for PhiX.
  makeudb_sintax  Make indexed database for sintax.
  orient Re-orient nt sequences to same strand as a reference database (useful for mixed-strand reads).
  search_phix  Search for PhiX.
  sintax  Taxonomy prediction by SINTAX algorithm.
  tree2clusters  Generate clusters from a tree by setting an identity cutoff.
  uchime2_denovo  Find chimeras in denoised amplicon sequences.
  uchime2_ref  Find chimeras by UCHIME2 algorithm using reference database.
  unoise  Denoising (error correction) of amplicon reads.

Backwards compatibility with v9
Otherwise, version 10 should be mostly backwards compatible with version 9 as there are relatively few changes to the internal code compared with previous major version updates. If users run into any compatibility problems I will update this page with details.
The derep_fulllength command is still supported but is deprecated. Use fastx_uniques for new scripts.

No longer supported
  uchime_denovo -- Does not work well with Illumina. Use an old version of usearch if you have 454 reads.
  uchime_ref -- Replaced by uchime2_ref.