Scan contig files against traditional PubMLST typing schemes
% mlst contigs.fa
contigs.fa neisseria 11149 abcZ(672) adk(3) aroE(4) fumC(3) gdh(8) pdhC(4) pgm(6)
% mlst genome.gbk.gz
genome.gbk.gz sepidermidis 184 arcC(16) aroE(1) gtr(2) mutS(1) pyrR(2) tpiA(1) yqiL(1)
% mlst --label Anthrax GCF_001941925.1_ASM194192v1_genomic.fna.gz
Anthrax bcereus - glp(24) gmk(1) ilv(~83) pta(1) pur(~71) pyc(37) tpi(41)
% mlst --nopath /opt/data/refseq/S_pyogenes/*.fna
NC_018936.fna spyogenes 28 gki(4) gtr(3) murI(4) mutS(4) recP(4) xpt(2) yqiL(4)
NC_017596.fna spyogenes 11 gki(2) gtr(6) murI(1) mutS(2) recP(2) xpt(2) yqiL(2)
NC_008022.fna spyogenes 55 gki(11) gtr(9) murI(1) mutS(9) recP(2) xpt(3) yqiL(4)
NC_006086.fna spyogenes 382 gki(5) gtr(52) murI(5) mutS(5) recP(5) xpt(4) yqiL(3)
NC_008024.fna spyogenes - gki(5) gtr(11) murI(8) mutS(5) recP(15?) xpt(2) yqiL(1)
NC_017040.fna spyogenes 172 gki(56) gtr(24) murI(39) mutS(7) recP(30) xpt(2) yqiL(33)
If you are using the OSX Brew or LinuxBrew packaging system:
% brew tap brewsci/science
% brew tap brewsci/bio
% brew untap homebrew/science
% brew install mlst # or "brew upgrade mlst"
If you are using Conda
% conda install -c bioconda -c conda-forge mlst
% cd $HOME
% git clone https://github.com/tseemann/mlst.git
% $HOME/mlst/bin/mlst --help
- NCBI BLAST+ blastn
- You probably have
blastnalready installed already. - If you use Brew, this will install the
blastpackage for you.
- You probably have
- Perl modules:
Moo,List::MoreUtils,JSON,File::Slurp- Debian:
sudo apt-get install libmoo-perl liblist-moreutils-perl libjson-perl libfile-slurp-perl - Redhat:
sudo apt-get install perl-Moo perl-List-MoreUtils perl-JSON perl-File-Slurp - Most Unix:
sudo cpan Moo List::MoreUtils JSON File::Slurp
- Debian:
- Standard Unix commands
file(for file format detection)gzip(for decompressing .gz files)
Simply just give it a genome file in FASTA or GenBank format, optionally compressed with gzip!
% mlst contigs.fa
contigs.fa neisseria 11149 abcZ(672) adk(3) aroE(4) fumC(3) gdh(8) pdhC(4) pgm(6)
It returns a tab-separated line containing
- the filename
- the matching PubMLST scheme name
- the ST (sequence type)
- the allele IDs
You can give it multiple files at once, and they can be in FASTA or GenBank format, and even compressed with gzip!
% mlst genomes/*
genomes/6008.fna saureus 239 arcc(2) aroe(3) glpf(1) gmk_(1) pta_(4) tpi_(4) yqil(3)
genomes/strep.fasta.gz ssuis 1 aroA(1) cpn60(1) dpr(1) gki(1) mutS(1) recA(1) thrA(1)
genomes/NC_002973.gbk lmonocytogenes 1 abcZ(3) bglA(1) cat(1) dapE(1) dat(3) ldh(1) lhkA(3)
genomes/L550.gbk.gz leptospira 152 glmU(26) pntA(30) sucA(28) tpiA(35) pfkB(39) mreA(29) caiB(29)
You can force a particular scheme (useful for reporting systems):
% mlst --scheme neisseria NM*
NM003.fa neisseria 4821 abcZ(222) adk(3) aroE(58) fumC(275) gdh(30) pdhC(5) pgm(255)
NM005.gbk neisseria 177 abcZ(7) adk(8) aroE(10) fumC(38) gdh(10) pdhC(1) pgm(20)
NM011.fa neisseria 11 abcZ(2) adk(3) aroE(4) fumC(3) gdh(8) pdhC(4) pgm(6)
NMC.gbk.gz neisseria 8 abcZ(2) adk(3) aroE(7) fumC(2) gdh(8) pdhC(5) pgm(2)
You can make mlst behave like older version before auto-detection existed
by providing the --legacy parameter with the --scheme parameter. In that case
it will print a fixed tabular output with a heading containing allele names specific to that scheme:
% mlst --legacy --scheme neisseria *.fa
FILE SCHEME ST abcZ adk aroE fumC gdh pdhC pgm
NM003.fa neisseria 11 2 3 4 3 8 4 6
NM009.fa neisseria 11149 672 3 4 3 8 4 6
MN043.fa neisseria 11 2 3 4 3 8 4 6
NM051.fa neisseria 11 2 3 4 3 8 4 6
NM099.fa neisseria 1287 2 3 4 17 8 4 6
NM110.fa neisseria 11 2 3 4 3 8 4 6
To see which PubMLST schemes are supported:
% mlst --list
abaumannii achromobacter aeromonas afumigatus cdifficile efaecium
hcinaedi hparasuis hpylori kpneumoniae leptospira
saureus xfastidiosa yersinia ypseudotuberculosis yruckeri
The above list is shortened. You can get more details using mlst --longlist.
achromobacter nusA rpoB eno gltB lepA nuoL nrdA
abaumannii Oxf_gltA Oxf_gyrB Oxf_gdhB Oxf_recA Oxf_cpn60 Oxf_gpi Oxf_rpoD
abaumannii_2 Pas_cpn60 Pas_fusA Pas_gltA Pas_pyrG Pas_recA Pas_rplB Pas_rpoB
aeromonas gyrB groL gltA metG ppsA recA
aphagocytophilum pheS glyA fumC mdh sucA dnaN atpA
arcobacter aspA atpA glnA gltA glyA pgm tkt
afumigatus ANX4 BGT1 CAT1 LIP MAT1_2 SODB ZRF2
bcereus glp gmk ilv pta pur pyc tpi
<snip>
Version 2.x does not just look for exact matches to full length alleles. It attempts to tell you as much as possible about what it found using the notation below:
| Symbol | Meaning | Length | Identity |
|---|---|---|---|
n |
exact intact allele | 100% | 100% |
~n |
novel full length allele similar to n | 100% | ≥ --minid |
n? |
partial match to known allele | ≥ --mincov |
≥ --minid |
- |
allele missing | < --mincov |
< --minid |
n,m |
multiple alleles |
Each MLST prediction gets a score out of 100. The score for a scheme with N alleles is as follows:
- +90/N points for an exact allele match e.g.
42 - +63/N points for a novel allele match (50% of an exact allele) e.g.
~42 - +18/N points for a partial allele match (20% of an exact alelle) e.g.
42? - 0 points for a missing allele e.g.
- - +10 points if there is a matching ST type for the allele combination
It is possible to filter results using the --minscore option which takes a
value between 1 and 100. If you only want to report known ST types, then use
--minscore 100. To also include novel combinations of existing alleles with
no ST type, use --minscore 90. The default is --minscore 50 which is an
ad hoc value I have found allows for genuine partial ST matches
but eliminates false positives.
The output is TSV (tab-separated values). This makes it easy to parse and manipulate with Unix utilities like cut and sort etc. For example, if you only want the filename and ST you can do the following:
% mlst --scheme abaumanii AB*.fasta | cut -f1,3 > ST.tsv
If you prefer CSV because it loads more smoothly into MS Excel, use the --csv option:
% mlst --csv Peptobismol.fna.gz > mlst.csv
JSON output is available too:
% mlst -q --json out.json --label Sepi test/example.fna.gz
% cat out.json
{
"Sepi" : {
"sequence_type" : "184",
"alleles" : {
"arcC" : "16",
"yqiL" : "1",
"gtr" : "2",
"pyrR" : "2",
"tpiA" : "1",
"aroE" : "1",
"mutS" : "1"
},
"filename" : "test/example.fna.gz",
"scheme" : "sepidermidis"
}
}
You can also save the "novel" alleles:
% mlst -q --novel who_knows.fa genome.fasta
Included is a file called db/scheme_species_map.tab which has 3
tab-separated columns as follows:
#SCHEME GENUS SPECIES
abaumannii Acinetobacter baumannii
abaumannii_2 Acinetobacter baumannii
achromobacter Achromobacter
aeromonas Aeromonas
afumigatus Aspergillus afumigatus
arcobacter Arcobacter
bburgdorferi Borrelia burgdorferi
bhampsonii Brachyspira hampsonii
bhenselae Bartonella henselae
borrelia Borrelia
bpilosicoli Brachyspira pilosicoli
<snip>
Note that that some schemes are species specific, and others are genus
specific, so the SPECIES column is empty. Note that the same
species/genus can apply to multiple schemes, see abaumanii above.
The mlst software comes bundled with the traditional MLST databases;
namely those schemes with less than 10 genes. I strive to make regular
releases with updated databases, but if this is not frequent enough you
can update the databases yourself using some tools included in the scripts
folder as follows:
# Figure out where mlst is installed
% which mlst
/home/user/sw/mlst
# Go into the scripts folder (you need to have write access!)
% cd /home/user/sw/mlst/scripts
# Run the downloader script (you need 'wget' installed)
% ./mlst-download_pub_mlst | bash
# Check it downloaded everything ok
% find pubmlst | less
# Save the old database folder
% mv ../db/pubmlst ../db/pubmlst.old
# Put the new folder there
% mv ./pubmlst ../db/
# Regenerate the BLAST database
% ./mlst-make_blast_db
# Check schemes are installed
% ../bin/mlst --list
If you are unable or unwilling to add your scheme to PubMLST via
BIGSdb you can
insert a new scheme into your local mlst database.
Each MLST scheme exists in a folder withing the mlst/db/pubmlst folder.
The name of the folder is the scheme name, say saureus for
Staphylococcus aureus. It contains files like this:
% cd mlst/db/pubmlst/sareus
% ls -1
saureus.txt
arcC.tfa
aroE.tfa
glpF.tfa
gmk.tfa
pta.tfa
saureus.txt
tpi.tfa
yqiL.tfa
The saureus.txt is a tab-separated file containing one ST definition
per row. The header line must be present. Extra columns with names
mlst_clade,clonal_complex,species,CC,Lineage are ignored.
% head -n 5 saureus.txt
ST arcC aroE glpF gmk pta tpi yqiL clonal_complex
1 1 1 1 1 1 1 1
2 2 2 2 2 2 2 26
3 1 1 1 9 1 1 12
4 10 10 8 6 10 3 2
Each of the .tfa files are nucleotide FASTA files with the allele
sequences for each locus. There must be a .tfa file for each and every
allele locus in the TSV scheme .txt file. Here is what the arcC.tfa
file looks like:
% head -n 20 arcC.tfa
>arcC_1
TTATTAATCCAACAAGCTAAATCGAACAGTGACACAACGCCGGCAATGCCATTGGATACT
TGTGGTGCAATGTCACAGGGTATGATAGGCTATTGGTTGGAAACTGAAATCAATCGCATT
TTAACTGAAATGAATAGTGATAGAACTGTAGGCACAATCGTTACACGTGTGGAAGTAGAT
AAAGATGATCCACGATTCAATAACCCAACCAAACCAATTGGTCCTTTTTATACGAAAGAA
GAAGTTGAAGAATTACAAAAAGAACAGCCAGACTCAGTCTTTAAAGAAGATGCAGGACGT
GGTTATAGAAAAGTAGTTGCGTCACCACTACCTCAATCTATACTAGAACACCAGTTAATT
CGAACTTTAGCAGACGGTAAAAATATTGTCATTGCATGCGGTGGTGGCGGTATTCCAGTT
ATAAAAAAAGAAAATACCTATGAAGGTGTTGAAGCG
>arcC_2
TTATTAATCCAACAAGCTAAATCGAACAGTGACACAACGCCGGCAATGCCATTGGATACT
TGTGGTGCAATGTCACAAGGTATGATAGGCTATTGGTTGGAAACTGAAATCAATCGCATT
TTAACTGAAATGAATAGTGATAGAACTGTAGGCACAATCGTAACACGTGTGGAAGTAGAT
AAAGATGATCCACGATTTGATAACCCAACTAAACCAATTGGTCCTTTTTATACGAAAGAA
GAAGTTGAAGAATTACAAAAAGAACAGCCAGGCTCAGTCTTTAAAGAAGATGCAGGACGT
GGTTATAGAAAAGTAGTTGCGTCACCACTACCTCAATCTATACTAGAACACCAGTTAATT
CGAACTTTAGCAGACGGTAAAAATATTGTCATTGCATGCGGTGGTGGCGGTATTCCAGTT
ATAAAAAAAGAAAATACCTATGAAGGTGTTGAAGCG
The FASTA sequence IDs must be named as >allele_number or
>allele-number. Ideally the sequences will not contain any
ambiguous IUPAC symbols. i.e. just A,T,C,G.
- Make a new folder in
mlst/db/pubmlst/SCHEME - Put your
SCHEME.txtfile in there - Put your
ALLELE.tfafiles in there - Run
mlst/scripts/mlst-make_blast_dbto update the BLAST indices - Run
mlst --longlist | grep SCHEMEto see if it exists - Run
mlst --scheme SCHEME file.fastato see if it works
If it doesn't - go back and check you really did do Step 4 above.
The mlst software incoprporates components of the
PubMLST database
and must be cited in any publications that use mlst:
"This publication made use of the PubMLST website (https://pubmlst.org/) developed by Keith Jolley (Jolley & Maiden 2010, BMC Bioinformatics, 11:595) and sited at the University of Oxford. The development of that website was funded by the Wellcome Trust".
You should also cite this software (currently unpublished) as:
- Seemann T,
mlstGithub https://github.com/tseemann/mlst
Please submit via the Github Issues page
- Torsten Seemann
- Web: https://tseemann.github.io/
- Twitter: @torstenseemann
- Blog: The Genome Factory