| Classifier | Version | Database built | Run | License |
|---|---|---|---|---|
| Blast | 2.7.1 | Yes | No | GPL2 |
| Centrifuge | 1.0.4_beta | Yes | No | GPL3 |
| Diamond | 0.9.24 | Yes | No | BSD |
| Kaiju | 1.6.3 | Yes | No | GPL3 |
| Kraken | 1.1 | Yes | No | GPL3 |
| Kraken | 2.0.7_beta | Yes | No | GPL3 |
| Krakenuniq | 0.5.7 | Yes | No | GPL3 |
| K-slam | 1.0 | Yes | No | GPL3 |
| Mmseqs2 | 8.fac81 | Yes | No | GPL3 |
| Paladin | 1.4.4 | Yes | No | MIT |
| Rapsearch | 2.24 | Yes | No | GPL3 |
| Salmon | 0.13.1 | No* | No | GPL3 |
| Sourmash | 2.0.0 | Yes | No | BSD |
* insufficient RAM
All the software have their docker image, using the same base and installed via conda
As of 2019-04-16 There are 13193 complete bacterial genomes, 8583 viral genomes and 283 archaeal genomes available in RefSeq complete genomes, for a total of 16877 genomes.
To download refseq, use make download, which will placed in the db directory
A summary, assemblies.csv will also be created with the columns "taxid", "species_taxid", "accession" and "organism_name"
To build the databases, use make build.
The databases are in db/$software_name
The following datasets are simulated using InSilicoSeq
| Abundance | Platform | N datasets | N reads |
|---|---|---|---|
| 10 | MiSeq, Novaseq | 5 | 5M |
| 50 | MiSeq, Novaseq | 5 | 5M |
| 200 | MiSeq, Novaseq | 5 | 5M |
for a total of 30 datasets. Genomes in theses datasets were selected at random from refseq.