Ensemble of HMM methods (SEPP, TIPP, UPP)
Switch branches/tags
Nothing to show
Clone or download
Latest commit 80c5274 Jun 20, 2018
Failed to load latest commit information.
sepp-package add decompose_tree symlink Jun 19, 2018
sepp fix identation issue Jun 20, 2018
test/unittest added unit tests for name collision detection Jun 14, 2018
tutorial Update tipp-tutorial.md Oct 19, 2017
.gitignore 4.2.1: May 8, 2017
.project Adding eclipse (pydev) .project file Dec 14, 2012
README.HIPPI.md fixed to say HIPPI instead of UPP in a couple places Feb 15, 2017
README.UPP.md 3.2.0: self-contained installation made possible Apr 2, 2017
README.md updated to new hmmer Sep 21, 2016
default.main.config by default do not pipe Oct 24, 2017
distribute_setup.py 3.0: update to python 3. first version Mar 27, 2017
merge_script.py update to python3 Mar 26, 2017
run_ensemble.py added new ability to just score sequences against an HMM family Jun 5, 2015
run_sepp.py new installation code and code structure Jan 18, 2013
setup.py updated version number and added changelog description Jun 14, 2018
split_sequences.py fix split_Sequences May 4, 2017



This repository includes code for SEPP, TIPP, UPP, HIPPI. The three methods use ensembles of Hidden Markov Models (HMMs) in different ways, each focusing on a different problem.

Each of these related tools has its own README file.


  • SEPP stands for "SATe-enabled Phylogenetic Placement", and addresses the problem of phylogenetic placement of short reads into reference alignments and trees.


  • TIPP stands for "Taxonomic Identification and Phylogenetic Profiling", and addresses the problem of taxonomic identification and abundance profiling of metagenomic data.


  • UPP stands for "Ultra-large alignments using Phylogeny-aware Profiles", and addresses the problem of alignment of very large datasets, potentially containing fragmentary data. UPP can align datasets with up to 1,000,000 sequences.


  • HIPPI stands for "Highly Accurate Protein Family Classification with Ensembles of HMMs", and addresses the problem of classifying query sequences to protein families.

Bugs and Errors

SEPP, TIPP, UPP, HIPPI are under active research development at UIUC by the Warnow Lab (and especially with her PhD student Mike Nute and former students Siavash Mirarab and Nam-phuong Nguyen). Please report any errors to Siavash Mirarab (smirarab@ucsd.edu), Nam Nguyen (ndn006@eng.ucsd.edu), or Mike Nute (nute2@illinois.edu).