Released: 2014-05-19
- A. What's CEGMA ?
- B. Installing CEGMA
- C. File Listing
- D. Compiling CEGMA
- E. To run CEGMA
- F. Authors and help
- G. Citing CEGMA
CEGMA (Core Eukaryotic Genes Mapping Approach) is a pipeline for building a set of high reliable set of gene annotations in virtually any eukaryotic genome. The strategy relies on a simple fact: some highly conserved proteins are encoded in essentially all eukaryotic genomes. We use the KOGs database to build a set of these highly conserved ubiquitous proteins. We define a set of 458 core proteins, and the protocol, CEGMA, to find orthologs of the core proteins in new genomes and to determine their exon-intron structures.
A local version of CEGMA can be installed on UNIX platforms and it requires pre-installation of Perl, NCBI BLAST+, HMMER, GeneWise and geneid. The procedure uses information from the core genes of six model organisms by first using TBLASTN to identify candidate regions in a new genome. It then proposes and redefines gene structures using a iterative combination of GeneWise, HMMER and geneid. The system includes the use of a profile for each core gene to ensure the reliability of the final predicted gene structure.
CEGMA source code, compiled binaries and documentation are available under the GNU GENERAL PUBLIC LICENSE.
The CEGMA distribution contains several directories and files. Source code and documentation files are included in the distribution.
The distribution is archived and compressed in a single file using the command tar -zcvf. The compressed file name is CEGMA.tar.gz (or something similar depending on compiled binaries included). The CEGMA files can be extracted following these instructions:
Type:
tar -zxvf cegma_v2.5.tar.gz
After executing these commands, the directory 'cegma_v2.5' will be created in your working directory.
CEGMA requires the pre-installation of the following software:
- geneid (geneid v1.4)
- genewise (wise2.2.3-rc7) or use our copy
- hmmer (HMMER v3.0)
- NCBI BLAST+ (tested using v2.2.29)
CEGMA has proved difficult to install on some Linux systems, these two guides that list how CEGMA was installed on Ubuntu might be therefore helpful:
To install Genewise, you will also need to have glib installed (which can be installed on a Mac via utilities such as Macports or Homebrew.
On an Ubuntu Linux system, you should also be able to install GeneWise by running:
sudo apt-get install wise
The CEGMA distribution contains the following files and directories:
bin/— The executable scriptsdata/— Core proteins, core profiles and cutoff and generic parameter file for geneidsample/— A DNA and protein file which can be used to test CEGMA.sample_output/— The results generated by CEGMA when using the included sample datasets.src/— Source code of CEGMA.GNULicense— This software is registered under GNU license.Makefile— This file is required to build CEGMA executable files.README.md— This file.
The CEGMA distribution contains a set of independent programs that are used by cegma.pl:
- parsewise - a parser for the genewise outputs.
+geneid-train and make_paramfile - build a parameter file for geneid
Change directory to the CEGMA directory and run the following to compile CEGMA:
make
This should generate the CEGMA executable files within the bin/ subdirectory. Now run:
cegma -h
This will test that the main CEGMA file can be executed.
There are two environmental variables that can be set by users to their preferences:
- You must specify the path where CEGMA can find the default files with the 'CEGMA' shell variable.
- CEGMA needs to write few temporary files in a directory with permissions for current user to read and write. Default temporary directory path is set to '/tmp' but you can assign a different temporary directory path using the variable 'CEGMATMP'.
- CEGMA uses some custom Perl modules (e.g. FAlite.pm, Cegma.pm). You must set the PERL5LIB environment variable to use CEGMA's 'lib' directory or you can copy the modules to another Perl module directory that is available to your Perl installation.
export CEGMA="path"
export CEGMATMP="path"
export PERL5LIB="$PERL5LIB:$CEGMA/lib"
setenv CEGMA "path"
setenv CEGMATMP "path"
setenv PERL5LIB "$PERL5LIB:$CEGMA/lib"
Genewise will also require that you set the $WISECONFIGDIR environment variable
To run CEGMA using the 458 default proteins type:
cegma --genome <genomic_fasta_sequence>
If you have multiple cores on your computer, you can speed things up by using the
-threads n option which passes the number of specified threads (specified by 'n')
to the TBLASTN and hmmsearch programs.
Run the following and compare the final output with the sample files sample_output/
cegma --genome sample.dna --protein sample.prot -o sample
CEGMA generates some intermediate files in the process. The files that contain the final predictions, in GFF and the fasta files of the corresponding genome and protein sequences are:
- sample.cegma.fa - predicted CEGs proteins
- sample.cegma.gff - coordinates in the genomic sequences
- sample.cegma.id - KOG ids for the selected proteins
- sample.cegma.local.dna - local fragments of DNA containing the genes
- sample.cegma.local.gff - coordinates in the local fragments
- sample.completeness_report - statistics of the percentages of CEGs
- sample.cegma.errors - may contain error messages produced by some programs
If you have a set of proteins that you want to use instead of the KOGs provided by CEGMA, you can do that easily. You have to create a HMM profile with HMMER, chose a cutoff for each profile and use the following options when running CEGMA:
-p, --protein fasta file of the protein sequences.
--prot_num Number of proteins per family/profile.
They have to be in consecutive order in the fasta file.
(default: 6)
--cutoff_file File with the cutoff for the HMMER alignments.
(default: \$CEGMA/data/profiles_cutoff.tbl)
--hmm_prefix Each protein ID must have "___" followed by the hmmprefix
and a number (ex. At3g02190___KOG1762).
(default: KOG)
--hmm_directory Directory that contains the hmm files. The files must be
named hmm_prefix(number).hmm ex. KOG1762.hmm.
(default: \$CEGMA/data/hmm_profiles)
Example:
cegma --genome sample.dna --prot_num 4 --protein ORTH.fa --hmm_prefix ORTH \
--hmm_profiles hmm_profiles/ --cutoff_file profiles_cutoff.tbl
For the previous command-line example, you must have 4 proteins per family and the proteins must be named protid___ORTH[0-1] (ex:At3g02190___ORTH0001). You must also have a directory with the hmm profile for each family name ORTH0001.hmm.
CEGMA has been written by Genis Parra (formerly at UC Davis Genome Center) and subsequently updated and maintained by Keith Bradnam (krbradnam@ucdavis.edu).
CEGMA home page is at http://korflab.ucdavis.edu/Datasets/cegma/
A FAQ with answers to common questions is also available: http://korflab.ucdavis.edu/Datasets/cegma/faq.html
CEGMA was first published in 2007 as a tool to train genefinders in novel genomes. In 2009 we adapted it to be able to estimate the completeness of the gene space in draft genomes, a measure often used as a proxy for assessing the completeness of a genome or transcriptome assembly. CEGMA can be cited by referring to these two papers:
- Genis Parra, Keith Bradnam and Ian Korf. CEGMA: a pipeline to accurately annotate core genes in eukaryotic genomes. Bioinformatics, 23: 1061-1067 (2007)
- Genis Parra, Keith Bradnam, Zemin Ning, Thomas Keane, and Ian Korf. Assessing the gene space in draft genomes. Nucleic Acids Research, 37(1): 298-297 (2009)