2010 Times Higher Education Award for University of the Year

 

SSuMMo Downloads

SSuMMo database from ARB108

This HMM database is trained from the latest ARB database release.
It contains profile hidden Markov models for all taxa represented in the ARB Silva database, release 108, trained using the default settings of hmmbuild.

md5 digest - aea93b35c1388710de4e7dbfbb3e5b2c

Minimised SSuMMo taxIndex for ARB108

This is the taxonomic index for SSUMMO release 108, with all training sequence accessions removed.
To update your existing SSuMMo installation, configure CONFIG.py to use this file instead of the old taxIndex.
Settings like `top' and `arbDBdir' need to be configured appropriately. Please see CONFIG.py for full details.

md5 digest - f636e044e74b03bdaad9013821bcbb82

SSuMMo taxIndex for ARB108

This contains the complete taxonomic index for SSUMMO release 108, with each training sequence accession assigned to the appropriate taxon.
This is unnecessary for SSuMMo usage, but is needed when building the HMMs with dictify.py.

md5 digest - cdd03822e9d5909784bbf2d05cb946b6

SSuMMo database from ARB104

This is the recommended database download.
It contains profile hidden Markov models for all taxa represented in the ARB Silva database, release 104, trained using the default settings of hmmbuild.

md5 digest - e24c6bceb48aeb09aee06f5c3ccd1f38

SSuMMo database from ARB104, with no prior sequence weights

This contains all the models as the above database, but these models were trained by giving the option '--wgiven' to hmmbuild. The difference is that by default, hmmbuild does some clever sequence weighting using prior probabilities. By giving the --wgiven option, hmmer does no prior weighting and so the HMMs here hold expected probabilities calculated directly from the observed training sequences.

MySQL database dumps

These are the mysql dumps for the taxonomy databases. They contain a minimised version of the NCBI taxonomy database, and tables which relate each taxon found in the ARB database to the NCBI taxonomic IDs and the ranks for each taxon, where found.

md5 digest - 6498426cbd9043f0d3fc15d5ae8875db