Usage notes
Pre-sketched databases are available for download below.
- Use the Primary links hosted at
http://faust.compbio.cs.cmu.eduif possible. We provide mirrors on google cloud, but this costs us more money. - See Parameter Guide below for algorithmic parameter information.
- See Database Descriptions below for database information.
.syl2db(two-stage) databases require sylph >= v1.0.0
Example usage:
# download database
wget http://faust.compbio.cs.cmu.edu/sylph-stuff/gtdb-r232-c200-dbv2.syl2db
# profile against database
sylph profile gtdb-r232-c200-dbv2.syl2db -1 sample_R1.fq -2 sample_R2.fq -t 30 > results.tsv
Note on taxonomy usage:
Most of the databases have associated taxonomies that sylph can utilize. See here for more information on taxonomy integration.
Databases (table scrolls to the right)
| Type | Name | Genomes | c-parameter | Size | Primary Download Link | Mirror | Notes |
|---|---|---|---|---|---|---|---|
| Prokaryotic (GTDB) | GTDB r232 (latest) | 199,923 species | -c 200, two-stage | 19 GB | gtdb-r232-c200-dbv2.syl2db | mirror | .syl2db, requires sylph >= v1.0.0 |
| GTDB r232 | 199,923 species | -c 200 | 24 GB | gtdb-r232-c200-dbv1.syldb | mirror | ||
| GTDB r232 | 199,923 species | -c 1000 | 4.9 GB | gtdb-r232-c1000-dbv1.syldb | mirror | ||
| GTDB r226 | 143,614 species | -c 200, two-stage | 14 GB | gtdb-r226-c200-dbv2.syl2db | mirror | .syl2db, requires sylph >= v1.0.0 |
|
| GTDB r226 | 143,614 species | -c 200 | 18 GB | gtdb-r226-c200-dbv1.syldb | mirror | ||
| GTDB r226 | 143,614 species | -c 1000 | 3.7 GB | gtdb-r226-c1000-dbv1.syldb | mirror | ||
| GTDB r220 | 113,104 species | -c 200 | 13.1 GB | gtdb-r220-c200-dbv1.syldb | mirror | ||
| GTDB r220 | 113,104 species | -c 1000 | 2.6 GB | gtdb-r220-c1000-dbv1.syldb | mirror | ||
| GTDB r214 | 85,202 species | -c 200 | 10 GB | v0.3-c200-gtdb-r214.syldb | mirror | ||
| GTDB r214 | 85,202 species | -c 1000 | 2 GB | v0.3-c1000-gtdb-r214.syldb | mirror | ||
| Prokaryotic (GlobDB) | GlobDB r232 (latest) | 346,233 species | -c 200, two-stage | 30 GB | globdb_r232_sylph_v2_c200.syl2db | .syl2db, requires sylph >= v1.0.0 |
|
| GlobDB r232 | 346,233 species | -c 200 | 37.5 GB | See the GlobDB website | Third-party database | ||
| GlobDB r232 | 346,233 species | -c 1000 | 7.5 GB | See the GlobDB website | Third-party database | ||
| GlobDB r226 | 306,260 species | -c 200 | 32 GB | See the GlobDB website | Third-party database | ||
| GlobDB r226 | 306,260 species | -c 1000 | 6.5 GB | See the GlobDB website | Third-party database | ||
| Prokaryotic (Other) | OceanDNA | 8,466 ocean MAGs | -c 200 | 800 MB | OceanDNA-c200-v0.3.syldb | mirror | |
| SMAG | 21,077 soil MAGs | -c 200 | 2.5 GB | SMAG-c200-v0.3.syldb | mirror | ||
| UHGG v2.0.1 (not dereplicated) | 289,232 gut genomes | -c 200 | 26 GB | uhgg_all_c200_v0.3.0.syldb | mirror | Not dereplicated - do not use for profiling | |
| Viral | UHGV | 171,338 gut vOTUs | -c 100 | 0.4 GB | uhgv_c100_dbv1.syldb | mirror | |
| UHGV | 171,338 gut vOTUs | -c 200 | 0.2 GB | uhgv_c200_dbv1.syldb | mirror | ||
| IMG/VR4.1 | 2,917,516 viral genomes | -c 200 | 2 GB | imgvr_c200_v0.3.0.syldb | mirror | ||
| Eukaryotic | RefSeq Fungi - latest | 661 genomes | -c 200 | 750 MB | fungi-refseq-2025-10-11-c200-v0.3.syldb | mirror | |
| TARA Oceans | 713 eukaryotic MAGs/SAGs | -c 200 | 900 MB | tara-eukmags-c200-v0.3.syldb | mirror |
Parameter Guide
- -c 200: More sensitive, larger file size
- -c 1000: More efficient, smaller file size, less sensitive
- -c 100: More sensitive but primarily for smaller genomes.
.syl2db(Preferred): New two-stage database format for genomes > 200 kbp. Requires sylph >= v1.0.0. Even faster than-c1000while having the sensitivity of-c200.
Note
-c 200 is used by default, so-c 100 must be specified for sketching if using a database with -c 100. For example:
# good
sylph sketch reads.fq -c 100 -d c100_reads/
sylph profile c100_database c100_reads/*
# the below does not work
# sylph sketch reads.fq -c 200 -d c200_reads/ # default
# sylph profile c100_database c200_reads/*
Database descriptions
GTDB Databases
The GTDB database is a high-quality, curated taxonomy and database for prokaryotes (archaea and bacteria). We take the dereplicated, species-representative genomes (one genome per species).
Available databases:
-
GTDB r232 database (199,923 species representative genomes) - April 18, 2026
-
GTDB r226 database (143,614 species representative genomes) - April 16, 2025
-
GTDB r220 database (113,104 species representative genomes) - April 24, 2024
-
GTDB r214 database (85,202 species representative genomes) - April 28, 2023
GlobDB - massive prokaryotic catalogue encompassing many other genome sets
GlobDB is a catalogue of > 300,000 prokaryotic genomes/MAGs. Their database is dereplicated at ~96% ANI. GlobDB encompasses 14 other large databases (including GTDB). The sylph database is hosted on their website.
Other prokaryotic databases
Tip
GlobDB encompasses SMAG and OceanDNA. We highly recommend using GlobDB if possible over these databases.
- OceanDNA catalogue of 8,466 ocean prokaryotic MAGs,
-c 200(800 MB) - SMAG catalogue of 21,077 soil MAGs,
-c 200(2.5 GB): - UHGG v2.0.1 catalogue of 289,232 gut genomes. Not dereplicated. Do not use for profiling.
-c 200(26 GB):
Viral databases
- Unified Human Gut Virome (UHGV) catalog - (171,338 vOTUs from human gut)
- Note: UHGV is more refined and has better taxonomic annotations than IMG/VR for human gut
- Pre-sketched IMG/VR4.1 database for high-confidence vOTU representatives (2,917,516 viral genomes).
Eukaryotic databases.
- 661 representative RefSeq fungi genomes (downloaded 2025-10-11),
-c 200 - 713 TARA Oceans eukaryotic MAGs/SAGs from Delmont et al.,
-c 200