Scripts to split reference and run mummer in parallel on an SGE cluster
Modified to run on 10x Genomics Cluster
INSTALL:
Download all three scripts:
git clone https://github.com/fritzsedlazeck/sge_mummer.git
Edit the main script: sge_mummer.sh
Open and correct the paths to the MUMmer installation and to the directory you placed all three scripts. Note that nucmer is called with the following parameters: " -maxmatch -c 100 ". This can be adjusted in: sge_mummer_helper.sh
Don't need any of the slurm files.
USAGE:
./sge_mummer.sh ref.fa query.fa tmp_directory
This script automates the process of "exploding" the reference multi-fasta file into a directory of individual fasta files, and then running in parallel MUMmer/nucmer of the full query dataset against each reference sequence. After the successful completion of the run the tmp_directory will include a merged_results.delta, which is the resulting file from the MUMmer alignments with the headers fixed to point to the original files. This delta file can be used as-if it had been computed directly from the full reference fasta file against the full query fasta file.
Questions or comments: fritz.sedlazeck@gmail.com