From https://github.com/combogenomics/medusa A draft genome scaffolder that uses multiple reference genomes in a graph-based approach. Thanks to EBosi and gianlucacolotto.
Requirements:
-
Mummer (sudo apt-get install mummer)
-
Minimap2 (sudo apt-get install minimap2)
-
Python > 3.6
-
Python modules: networkx, biopython, flask
medusa2web: contains medusa2web and script for running medusa2. Replace with your path on launcher.sh and start python3 medusa2web.py (http://127.0.0.1:5000/)
medusa2standalone: contains medusa2 standalone version for running directly from bash
Run & Required parameters: python3 medusa.py -i input_filename -f reference_folder
Optional parameters:
-s skipmap_folder (folder that contains coords or paf file generated previously from mummer or minimap2, with these you can skip alignment step)
-a use Minimap2 aligner instead of Mummer aligner
-t <n> number of processes used for alignment step
-v <n> verbosity level
-o output folder for support files and scaffold writing
Under folder /medusa2standalone/BenchmarkMedusa2/ you can see many usage example for running the tool! In dataset_medusa.txt you have a link for dataset used about BCEN,ECOL,MTUB,RSPH,SAUR,THAL,CELEGANS.