In version 4.1 of the public-domain version of UCHIME, sequences in the query and reference database file MUST match case (i.e, query and reference sequences must all be upper-case or all lower-case), otherwise chimera detection will fail with no warning. This bug is fixed in v4.2.
The public-domain version of UCHIME does not support the --self option. Instead you should use the --selfid option, which ignores reference sequences that are 100% identical to the query. This is intended for use in benchmarking for measuring false positives using a database of trusted chimera-free sequences.
|USEARCH||Includes an implementation of the UCHIME algorithm that is highly optimized for speed.|
|Reference databases||A good reference database for 16S ribosomal RNA genes is available in the Microbiome Utilities provided by the Broad Institute. An alternative could be the chimera.slayer reference database provided for the use in the mothur implementation of an algorithm very similar to ChimeraSlayer.|
|"Gold" database||This is a FASTA file containing the ChimeraSlayer reference database in the Broad Microbiome Utilities (see above), version microbiomeutil-r20110519. This contains all sequences in both orientations (forward and reverse-complemented).|
|uchime4.2.40_src.tar.gz||Source code for public domain version of UCHIME.|
|uchime4.2.40_linuxi86||Precompiled Linux binary (32-bit i86) built from the public-domain source code.|
|chsim1.0.tar.gz||CHSIM source code, precompiled binaries and documentation.|
|simm.tar.gz||SIMM dataset (FASTA files).|