Contents

• What is RiboGrove
  • RiboGrove and other 16S rRNA databases
  • Genome categories
• Downloads
  • Latest RiboGrove release — 27.233
  • RiboGrove release archive
  • Release notes
• Statistical summary
  • RiboGrove size
  • 16S rRNA gene lengths
  • 16S rRNA gene copy number
  • Top-10 longest 16S rRNA genes
  • Top-10 shortest 16S rRNA genes
  • Top-10 genomes with the largest 16S rRNA copy numbers
  • Top-10 genomes with the highest intragenomic variability of 16S rRNA genes
  • Coverage of primer pairs for different V-regions of bacterial 16S rRNA genes
• Searching data in RiboGrove
  • Header format
  • Search sequences by header
  • Search sequences by length
  • Selecting header data
• Contacts
• Citing RiboGrove
• Questions people ask about RiboGrove

What is RiboGrove

RiboGrove is a database of 16S rRNA gene sequences of bacteria and archaea.

RiboGrove is based on the RefSeq database. It contains only full-length sequences of 16S rRNA genes, and the sequences are derived from completely assembled prokaryotic genomes deposited in RefSeq. Hence we posit high reliability of RiboGrove sequences.

RiboGrove and other 16S rRNA databases

Here is a summary showing what is the (qualitative) difference between RiboGrove and similar rRNA sequence databases, namely rrnDB, Silva, RDP, and Greengenes. Briefly, RiboGrove is inferior in sequence amount and diversity, but superior in sequence reliability.

Genome categories

All genomes used for RiboGrove construction were divided into three categories according to their expected reliability:

1. Category 1 (the highest reliability). Genomes showing no signs of a low-quality assembly and sequenced either with PacBio technology or with a combination “Oxford Nanopore + Illumina”.
2. Category 2. Genomes showing no signs of a low-quality assembly and sequenced with any other technology (or the technology is not specified).
3. Category 3 (the lowest reliability). Genomes showing at least one sign of a low-quality assembly.

Signs of a low-quality assembly are the following:

• The genome contains degenerate base(s) in 16S rRNA gene sequences.
• The assembly includes at least one RefSeq record whose title contains the phrase “map unlocalized” and this record contains a 16S rRNA gene or a part of it.

The software used for the RiboGrove construction can be found in the following GitHub repository: ribogrove-tools.

Downloads

Latest RiboGrove release — 27.233 (2026-02-03)

The release is based on RefSeq release 233.

DOI of RiboGrove release 27.233: 10.5281/zenodo.17273649.

• A fasta file of full-length 16S gene sequences. Download (gzipped fasta file, 11.24 MB)
• Metadata Download (zip archive 7.43 MB)
  
  What information exactly does the metadata contain?

  The metadata consists of the following files:

  1. discarded_sequences.fasta.gz
     This is a fasta file of sequences, which were present in source RefSeq genomes and were annotated a 16S rRNA genes but which have been discarded according to their incompleteness, internal repeats etc. and thus haven’t been included into RiboGrove.
  2. source_RefSeq_genomes.tsv
     This is a TSV files, which contains information about what genomes were used for the RiboGrove construction.
  3. gene_seqs_base_counts.tsv, discarded_gene_seqs_base_counts.tsv
     These are TSV files, which contain nucleotide composition and size of the gene sequences. The first file describes final RiboGrove gene sequences, the second file describes discarded sequences.
  4. categories.tsv
     This is a TSV file, which contains information about what genome categories were assigned to each genome and why. Moreover, it contains information about what sequencing technology was used to sequence each genome.
  5. taxonomy.tsv
     This is a TSV file, which contains taxonomic affiliation of each genome and gene.
  6. intragenic_repeats.tsv
     This is a TSV file, which contains information about intragenomic repeats found in gene sequences using RepeatFinder.
  7. entropy_summary.tsv
     This is a TSV file, which contains summary of instragenomic variability of the 16S rRNA genes. Intragenomic variability are calculated only for the category 1 genomes having more than one 16S rRNA gene. Intragenomic variability is evaluated using Shannon entropy. We align gene sequences from each genome using MAFFT, and then we calculate Shannon entropy for each multiple alignment column (i.e. base).
  8. 16S_GCNs.tsv
     This is a TSV file of 16S rRNA Gene Copy Numbers for each genome in the release.
  9. primer_pair_genomic_coverage.tsv
     This is a TSV file which contains genomic coverage of primer pairs targeting different V-regions of 16S rRNA genes. For example, for Enterobacteriaceae, genomic coverage of a primer pair is the percent of Enterobacteriaceae genomes which contain at least one 16S rRNA gene that can (theoretically) produce a PCR product using the primer pair.

The fasta file is compressed with gzip, and the metadata file is a zip archive. To uncompress them, Linux and Mac OS users may use gzip and zip programs, they should be built-in. For Windows users, the free and open-source (de)compression program 7-Zip is available.

RiboGrove release archive

You can find all releases in the RiboGrove release archive.

Release notes

No important differences from the previous release.

You can find notes to all RiboGrove releases on the release notes page.

Statistical summary

^* Metrics marked with an asterisk were calculated with preliminary normalization, i.e. median within-species gene length was used for the summary.

^* These are median within-species copy numbers.

^* Entropy is Shannon entropy calculated for each column of the multiple sequence alignment (MSA) of all full-length 16S rRNA genes of a genome. Entropy is then summed up (column “Sum of entropy”) and averaged (column “Mean entropy”).

^** Halomicrobium sp. ZPS1 is a quite remarkable case. This genome harbours two 16S rRNA genes, therefore entropy is equal to the number of mismatching nucleotides between sequences of the genes. Respectively, percent of identity between these two gene sequences is 90.70%! This is remarkable because the usual (however arbitrary) genus demarcation threshold of percent of identity is 95%.

Coverage^* of primer pairs for different V-regions of 16S rRNA genes

^* Coverage of a primer pair is the percent of genomes having at least one 16S rRNA gene which can be amplified by PCR using this primer pair. For details, see our paper about RiboGrove.

In the tables below, you can find coverage of primer pairs that are being commonly used to amplify bacterial and archaeal genes (“bacterial” and “archaeal” primers).

You can find a more detailed table in the file primer_pair_genomic_coverage.tsv in the metadata. That table contains coverage not just for phyla, but also for each kingdom, class, order, family, genus, and species. Moreover, that table contains coverage values for additional primer pairs, namely 1115F-1492R, 349f-519r, 1106F-Ar1378R, 1106F-SSU1492Rngs, SSU1ArF-SSU468R, SSU1ArF-SSU520R. In the tables below, they are omitted for brevity.

Searching data in RiboGrove

RiboGrove is a very minimalistic database — it comprises a collection of plain fasta files with metadata. Thus, extended search instruments are not available for it. We admit this problem and provide a list of suggestions below. The suggestions would help you to explore and select RiboGrove data.

Header format

RiboGrove fasta data has the following format of header:

>GCF_000978375.1:NZ_CP009686.1:8908-10459:plus ;d__Bacteria;k__Bacillati;p__Bacillota;c__Bacilli;o__Bacillales;f__Bacillaceae;g__Bacillus;s__cereus; category:1

Major blocks of a header are separated by spaces. A header consists of three such blocks:

1. Sequence ID (seqID): GCF_000978375.1:NZ_CP009686.1:8908-10459:plus. SeqID, in turn, consists of four parts separated by semicolons (:):
   1. The Assembly accession of the genome from which the gene originates: GCF_000978375.1.
   2. The accession number of the RefSeq sequence, from which the gene originates: NZ_CP009686.1.
   3. Coordinates of the gene within this RefSeq genomic sequence: 8908-10459 (coordinates are 1-based, left-closed and right-closed).
   4. Strand of the RefSeq genomic sequence, where the gene is located: plus (or minus).
2. A taxonomy string, comprising domain (Bacteria), kingdom (Bacillati), phylum (Bacillota), class (Bacilli), order (Bacillales), family (Bacillaceae), genus (Bacillus) names, and the specific epithet (cereus).
   Each name is preceded by a prefix, which denotes rank: d__ for domain, k__ for kingdom, p__ for phylum, c__ for class, o__ for order, f__ for family, g__ for genus, and s__ for specific epithet. Prefixes contain double underscores.
   The taxonomic names are separated and flanked by semicolons (;).
3. The category of the genome, from which the gene sequence originates: (category:1).

Sequence selection

You can select specific sequences from fasta files using the Seqkit program (GitHub repo, documentation). It is free, cross-platform, multifunctional and pretty fast and can process both gzipped and uncompressed fasta files. Programs seqkit grep and seqkit seq are useful for sequence selection.

Search sequences by header

Given the downloaded fasta file ribogrove_27.233_sequences.fasta.gz, consider the following examples of sequence selection using seqkit grep:

Example 1. Select a single sequence by SeqID.

seqkit grep -p "GCF_000978375.1:NZ_CP009686.1:8908-10459:plus" ribogrove_27.233_sequences.fasta.gz

The -p option sets a pattern to search in fasta headers (only in sequence IDs, actually).

Example 2. Select all gene sequences of a single RefSeq genomic sequence by accession number NZ_CP009686.1.

seqkit grep -nrp ":NZ_CP009686.1:" ribogrove_27.233_sequences.fasta.gz

Here, two more options are required: -n and -r. The former tells the program to match the whole headers instead of IDs only. The latter tells the program to include partial matches into output, i.e. if the pattern is a substring of a header, the header will be printed to output.

To ensure search specificity, surround the Accession.Version with colons (:).

Example 3. Select all gene sequences of a single genome (Assembly accession GCF_019357495.1).

seqkit grep -nrp "GCF_019357495.1:" ribogrove_27.233_sequences.fasta.gz

To ensure search specificity, put a colon (:) after the assembly accession.

Example 4. Select all actinobacterial sequences.

seqkit grep -nrp ";p__Actinobacteria;" ribogrove_27.233_sequences.fasta.gz

To ensure search specificity, surround the taxonomy name with semicolons (;).

Example 5. Select all sequences originating from category 1 genomes.

seqkit grep -nrp "category:1" ribogrove_27.233_sequences.fasta.gz

Example 6. Select all sequences except for those belonging to Bacillota.

seqkit grep -nvrp ";p__Bacillota;" ribogrove_27.233_sequences.fasta.gz

Recognize the -v option within the option sequence -nvrp. This option inverts match, i.e. output will comprise sequences, headers of which do not contain the substring “;p__Bacillota;”.

Search sequences by length

You can use the seqkit seq program to select sequences by length.

Example 1. Select all sequences longer than 1600 bp.

seqkit seq -m 1601 ribogrove_27.233_sequences.fasta.gz

The -m option sets the minimum length of a sequence to be printed to output.

Example 2. Select all sequences shorter than 1500 bp.

seqkit seq -M 1499 ribogrove_27.233_sequences.fasta.gz

The -M option sets the maximum length of a sequence to be printed to output.

Example 3. Select all sequences having length in range [1500, 1600] bp.

seqkit seq -m 1500 -M 1600 ribogrove_27.233_sequences.fasta.gz

Selecting header data

It is sometimes useful to retrieve only header information from a fasta file. You can use the seqkit seq program for it.

Example 1. Select all headers.

seqkit seq -n ribogrove_27.233_sequences.fasta.gz

The -n option tells the program to output only headers.

Example 2. Select all SeqIDs (header parts before the first space).

seqkit seq -ni ribogrove_27.233_sequences.fasta.gz

The -i option tells the program to output only sequence IDs.

Example 3. Select all RefSeq “Assession.Version”s.

seqkit seq -ni ribogrove_27.233_sequences.fasta.gz | cut -f2 -d':' | sort | uniq

This might be done only if you have cut, sort and uniq utilities installed (Linux and Mac OS systems should have them built-in).

Example 4. Select all Assembly accessions.

seqkit seq -ni ribogrove_27.233_sequences.fasta.gz | cut -f1 -d':' | sort | uniq

This might be done only if you have cut, sort and uniq utilities installed (Linux and Mac OS systems should have them built-in).

Example 5. Select all phylum names.

seqkit seq -n ribogrove_27.233_sequences.fasta.gz | grep -Eo ';p__[^;]+' | sed -E 's/;|p__//g' | sort | uniq

This might be done only if you have grep, sed, sort and uniq utilities installed (Linux and Mac OS systems should have them built-in).

Contacts

For any questions concerning RiboGrove, please contact Maksim Sikolenko at sikolenko@mbio.bas-net.by or maximdeynonih@gmail.com.

Citing RiboGrove

If you find RiboGrove useful for your research please cite:

Maxim A. Sikolenko, Leonid N. Valentovich. “RiboGrove: a database of full-length prokaryotic 16S rRNA genes derived from completely assembled genomes” // Research in Microbiology, Volume 173, Issue 4, May 2022, 103936.
(DOI: 10.1016/j.resmic.2022.103936).

You can also cite RiboGrove itself on Zenodo:

• RiboGrove of no particular release: DOI: 10.5281/zenodo.17190266;
• RiboGrove release 17273649 specifically: DOI: 10.5281/zenodo.17273649.

Questions people ask about RiboGrove

RiboGrove, 2026-02-03