Other databases

NanoClass was designed to classify 16S or 18S rRNA gene amplicon sequences based on the SILVA database. Most tools implemented in NanoClass2 will be able to classify sequences using alternative databases. The use of custom databases has not been fully tested and comes with no warranties, but some advice on how to prepare alternative databases, such as Unite, can be found here.

Important

To reduce runtime, NanoClass2 will only run those steps in the pipeline for which no output files are detected. This means that if a database was downloaded and build during an earlier run of NanoClass2, it will automatically use this database instead of downloading and building a new one. This saves a lot of time and computational resources, but also requires care by the user to make sure the intended database is used.

It is recommended to use a fresh NanoClass2 application if a different database is used. Simply clone or download NanoClass2 again following the instructions in the quick start section of this documentation.

Custom databases

Most classification tools implemented in NanoClass2 can use alternative databases supplied by the user (see exceptions below), provided they are formatted correctly. These databases will not be automatically downloaded as part of the NanoClass pipeline, as is the case for the SILVA 16S or 18S databases. Instead, users should provide the databases themselves and store them in the db/common/ subdirectory of the NanoClass2 directory. These will then be automatically detected once NanoClass2 is started and NanoClass2 will create and reformat all tools-specific databases based on this user-provided database.

Two files are required:

The reference sequences

The reference nucleotide sequences should be in an unzipped, multi-fasta file called ref-seqs.fna. The fasta header should have the sequence identifier only.

Here is an example of an entry in the ref-seqs.fna of the default SILVA reference database:

>AY846379.1.1791
AACCTGGTTGATCCTGCCAGTAGTCATATGCTTGTCTCAAAGATTAAGCCATGCATGTCTAAGTATAAACTGCTTATACT
GTGAAACTGCGAATGGCTCATTAAATCAGTTATAGTTTATTTGATGGTACCTCTACACGGATAACCGTAGTAATTCTAGA

The reference taxonomy

The reference taxonomy should be in an unzipped tab-delimited file called ref-taxonomy.txt. The first column contains the sequence identifier corresponding to the identifiers used in the ref-seqs.fna. The second column contains a “;”-separated 7-level taxonomy string with domain;phylum;class;order;family;genus;species.

Here is an example of some of the entries in ref-taxonomy.txt of the default SILVA reference database.

LT595716.1.1435 D_0__Bacteria;D_1__Firmicutes;D_2__Bacilli;D_3__Lactobacillales;D_4__Carnobacteriaceae;D_5__Jeotgalibaca;D_6__NA
AB696431.1.1480 D_0__Bacteria;D_1__Firmicutes;D_2__Bacilli;D_3__uncultured;D_4__NA;D_5__NA;D_6__NA
AACY020207233.1273.2814 D_0__Bacteria;D_1__Marinimicrobia (SAR406 clade);D_2__NA;D_3__NA;D_4__NA;D_5__NA;D_6__NA

Example using Unite

Although not automatically implemented in NanoClass2, it is possible to manually download the ITS UNITE database to be used in NanoClass2. Note, that this might only work for a subset of databases.

First download the UNITE database from here. Then unzip it and look for the reference sequences (sh_refs_qiime_ver8_99_04.02.2020.fasta) and the reference taxonomy (sh_taxonomy_qiime_ver8_99_04.02.2020.txt).

Move these two files to the to a db/common/ directory in the NanoClass2 folder and rename them to ref-seqs.fna and ref-taxonomy.txt, respectively.

mkdir -p db/common/
mkdir -p db/common/taxonomy
mkdir -p db/common/data

mv sh_refs_qiime_ver8_99_04.02.2020.fasta db/common/ref-seqs.fna
mv sh_taxonomy_qiime_ver8_99_04.02.2020.txt db/common/ref-taxonomy.txt

Lastly, you need to create some empty files.

#dummy file for mothur
touch db/common/ref-seqs.aln

#dummy files for kraken and centrifuge
touch db/common/seqid2taxid.map
touch db/common/taxonomy/names.dmp
touch db/common/taxonomy/nodes.dmp
touch db/common/data/SILVA_138.1_SSURef_NR99_tax_silva.fasta
touch db/common/ref-seqs-tax.fna

These empty files are expected by some classifiers and you might run into issues when not creating these dummy files first.

For now, these steps should allow you to use the following classifiers: blastn, dcmegablast, idtaxa, megablast, minimap, qiime, rdp and spingo.

Exceptions

Mothur

For Mothur an additional fasta file is required with the multiple sequence alignments of all reference sequences in the database. The fasta header should be the same as the fasta header of the sequence file. The file should be named ref-seqs.aln and stored in db/common in the NanoClass2 folder.

If you do not have an alignment for your custom database, you can simply skip mother in the NanoClass2 run by removing it from the methods list in the config.yaml

Kraken2

Kraken2 cannot be used within NanoClass in combination with custom databases at this stage.

Centrifuge

Centrifuge cannot be used within NanoClass in combination with custom databases at this stage.