ECemble

A hierarchical machine learning method for enzyme class prediction

ECemble is an open source ensemble machine learning pipeline tool (released under the GNU General Public License v3) that allow users to efficiently and automatically process proteomes to predict enzyme and enzyme classes from unannotated protein sequences. ECemble uses various learning algorithms to generate multiple prediction models that distinguish different classes of enzymes, where it first predicts if a protein is an enzyme or a non-enzyme, and then subsequently predict specific class and subclass of an enzyme in the EC number hierarchy. The predictions are selected when at least two of the three top-performing ML classifiers show consistent predictions.

Citation: If you use this resource, please cite the following reference: Mohammed A, Guda C. Application of Hierarchical enzyme classification method reveals the role gut microbiome in human metabolism. BMC Genomics 2015 https://bmcgenomics.biomedcentral.com/articles/10.1186/1471-2164-16-S7-S16

Note: ECemble is an open-source software, in case if you run across bugs or errors, raise an issue over here.

Table of Contents

This README file will serve as a guide for using this software tool. We suggest reading through the entire document at least once, in order to get an idea of the options available, and how to customize the pipeline to fit your needs.

Downloading ECemble

Clone the git repository:

$ git clone https://github.com/akram-mohammed/ECemble.git && cd ECemble

Dependencies

After downloading ECemble, make sure you install all the necessary software packages.

cd ECemble
mkdir bin/lib/PFAM bin/lib/PROSITE bin/lib/SUPERFAMILY bin/lib/WEKA bin/lib/hmmer

Installing PFAM

cd bin/lib/PFAM
wget ftp://ftp.ebi.ac.uk/pub/databases/Pfam/releases/Pfam26.0/Pfam-A.hmm.gz
gunzip Pfam-A.hmm.gz

Installing HAMMER

cd bin/lib/hmmer
wget ftp://selab.janelia.org/pub/software/hmmer3/3.1b1/hmmer-3.1b1-linux-intel-x86_64.tar.gz
tar zxf hmmer-3.1b1-linux-intel-x86_64.tar.gz
rm hmmer-3.1b1-linux-intel-x86_64.tar.gz
./configure
make 
make check
./src/hmmpress ../PFAM/Pfam-A.hmm

Installing PROSITE

cd bin/lib/PROSITE
wget ftp://ftp.expasy.org/databases/prosite/ps_scan/ps_scan_linux_x86_elf.tar.gz
tar zxf ps_scan_linux_x86_elf.tar.gz
rm ps_scan_linux_x86_elf.tar.gz
wget ftp://ftp.expasy.org/databases/prosite/prosite.dat

Installing SCOP

cd bin/lib/SUPERFAMILY
ftp supfam.org
username: license
password: SlithyToves
cd models
get model.tab.gz
get hmmlib_1.75.gz
get self_hits.tab.gz
cd ../sequences
get pdbj95d.gz
cd ../scripts
mget *
bye

wget http://scop.mrc-lmb.cam.ac.uk/scop/parse/dir.des.scop.txt_1.75
wget http://scop.mrc-lmb.cam.ac.uk/scop/parse/dir.cla.scop.txt_1.75
mv dir.des.scop.txt_1.75 dir.des.scop.txt
mv dir.cla.scop.txt_1.75 dir.cla.scop.txt

gunzip pdbj95d.gz
gunzip model.tab.gz
gunzip hmmlib_1.75.gz
mv hmmlib_1.75 hmmlib
gunzip self_hits.tab.gz
chmod u+x *.pl
.././hmmer/src/hmmpress hmmlib

#Change the paths in ass3.pl, line 13-16

my $selfhits = "./../lib/SUPERFAMILY/self_hits.tab";
my $clafile      = "./../lib/SUPERFAMILY/dir.cla.scop.txt";
my $modeltab     = "./../lib/SUPERFAMILY/model.tab";
my $pdbj95d      = "./../lib/SUPERFAMILY/pdbj95d";

#Make changes in superfamily.pl

line 20	system "perl ../lib/SUPERFAMILY/fasta_checker.pl $ARGV[0] >../test/scratch/$file\_torun.fa";
line 24 system "../lib/hmmer/src/hmmscan -o ../test/scratch/$file.res -E 1e-04 -Z 15438 ../lib/SUPERFAMILY/hmmlib ../test/scratch/$file\_torun.fa";
line 28 system "perl ../lib/SUPERFAMILY/ass3.pl -t n -f 40 -e 0.0001 ../test/scratch/$file\_torun.fa ../test/scratch/$file.res ../test/scratch/$file.ass ";
#comment lines
line 31 print "Running ass_to_html\n";
line 32 system "ass_to_html.pl dir.des.scop.txt model.tab $file.ass > $file.html";

Installing WEKA

wget http://prdownloads.sourceforge.net/weka/weka-3-7-6.zip
unzip weka-3-7-6.zip
mv weka-3-7-6 ../lib/
rm weka-3-7-6.zip
cd ../lib/weka-3-7-6
jar -xvf weka.jar
jar -xvf weka-src.jar

Next, set the WEKA classpath by entering the following command in .bashrc file under Alias definitions:

export WEKAINSTALL=/absolute/path/to/weka/directory/`
export CLASSPATH=$CLASSPATH:$WEKAINSTALL/weka.jar

To install ECemble dependencies right from scratch, check out our exhaustive guides:

System Requirements

You will need current or very recent generations of your operating system: Linux OS, Mac OS X.

Directory Structure of the Pipeline

After downloading ECemble, notice inside the ECemble directory there are all scripts necessary to process data:

Execution of Pipeline

The downloaded scripts are used to extract features, train models for each of the 4 EC levels and test models. Please create a ticket under issues section to ask any question related to execution of any script.

Team


Dr. Akram Mohammed
akrammohd@gmail.com

Author and Maintainer
Dr. Babu Guda
babu.guda@unmc.edu

Principal Investigator

License

This software has been released under the GNU General Public License v3.