Developing a software package to analysis metagenomics data

Language: Python 2.7 Required package

• Biopython - 1.64
• NumPy - 1.8.2
• SciPy - 0.12.1

Required software

• Genovo v0.4: http://cs.stanford.edu/group/genovo/
• Xgenovo http://xgenovo.dna.bio.keio.ac.jp/
• Glimmer v3.02: https://ccb.jhu.edu/software/glimmer/
  • requires ELPH v1.0.1: http://cbcb.umd.edu/software/ELPH/
• MINE v1.0.1: http://www.exploredata.net/
  • requires Java: http://www.java.com/
• NCBI BLAST+ v2.2.28+: http://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE=BlastDocs&DOC_TYPE=Download

Note: Binary files distributed here might NOT work with your OS!

The following folder structure is used in the default control file. If Genovo/Glimmer/MINE are located at other destinations, please update the control file.

• data
• example ## contain basic examples
• Genovo
• Glimmer
• MINE
• BLAST # Contains blastx, which is a symlink to /usr/bin/blastx
• src
• core
• test

• The control file in MMAP have two purpose. The main purpose is allowing MMAP to locate other software (genovo, glimmer, MINE, and blast).
• The secondary purpose is to allow users to customize the parameters used in the pipeline. Please refer to the Additional parameters section.

1. Genovo shipped with precompiled binaries. Check and run the commented demo script DEMO.sh.
2. If these precompiled binaries or the demo script fail. Recompile it from the src folder. The default Makefile requires libtool
3. Update control file accordingly. Make sure genovo_pdir points to the folder contains the following binaries assemble and finalize.

1. Compile Xgenovo.
2. Update control file accordingly. Make user 'xgenovo_pdir' points to the folder contains the following binaries assemble and finalize.

1. Glimmer often required some custom setup. Try to use full/absolute path if relative path doesn't work.
2. Follow the instruction in glim302notes.pdf and compile glimmer from the source code.
3. Download and install elph
4. Update awkpath, glimmerpath, and elphbin in glimmer/scritps/g3-iterated.csh * The awkpath should be points to the scritps folder * The glimmerpath should point to the bin folder * The elphbin is the full path points to the executable elph. Not the folder.
5. Update control file accordingly. Make sure glimmer_pdir points to the top level of the glimmer folder. MMAP is looking for bin and scripts underneath this folder.

The BLAST component queries annotated Gene Ontology sequences downloaded from http://geneontology.org. Prior to running an analysis, you must download the sequences and convert them to a BLAST+ formatted database.

A bash script is provided that does this automatically: [src/scripts/makeblastdb-go.sh]. It uses curl and makeblastdb to download a FASTA-formatted protein sequence and convert it to NCBI BLAST+ format.

It accepts one argument - the output directory for the GO Sequence Database.

$ ./makeblastdb-go.sh /data/go-seqdb

After generating the local database, add its path to the control file, so that the pipeline knows where to find it.

1. Install and make sure your java version is at least 1.7 (java -version)
2. Update control file accordingly. Make sure mine_pdir points to the folder contains MINE.jar.

• Update control file [default: MMAP/control]
• program_pdir points to the directory contains the exe
• blasd_db pointst to the local BLAST database

python MMAP/main.py -h
python MMAP/main.py summary -h
python MMAP/main.py process -h


## To run Genovo/Glimmer/Blast, use -i to provide input fasta file
python MMAP/main.py process -c data/example/control -i data/example/MMAP_example.fasta
python MMAP/main.py process -c data/example/controlX -i data/example/MMAP_paired_example.fasta

## To run MINE, use -m to provide a directory with list of csv files
python MMAP/main.py summary -m data/example/

## Custom control file can be used wit -c
python MMAP/main.py process -i data/example/MMAP_example.fasta -c path_to_custom_control_file