NanoPoor

Monitor tool for nanopore sequencing runs. It is used to estimate the long read coverage on a short reads assembly while sequencing

WARNING

THIS IS THE ALPHA VERSION. It is published as it was used for the first test. A beta version will soon be released

Dependencies

• Python3.6
• Deepnano-blitz (it is run in a conda environment with Python 3.6 and the Rust compiler)
• blast+
• biopython
• numpy

How to run NanoPoor

conda activate deepnanoblitz

python NanoPoor.py {folder_name} {reference_assembly}

the {folder_name} is the same as the one set in the MinKNOW software

{reference_assembly} needs to be in fasta format

How it works

• NanoPoor uses deepnano-blitz to perform a quick-and-dirty basecall

• the output fasta file is used as query in a blast-search against the short-read assembly

• this process is run every few minutes (default: every hour) on newly generated fast5 files and updates a dictionary containing the covarege for each base in the draft assembly

• NanoPoor performance is optimal when run using 10+ threads in a RAMdisk folder

Some issues still need to be addressed:

• needs a better implementation of the deepnano-blitz basecaller
• needs parameter settings and a help menu
• needs a graphical output
• needs to automatically detect an ongoing nanopore run