polish
Polishing nanopore reads is the process of taking draft genomes and improving them through additional comparisons of the draft genomes with the relevant reads. It is recommended to polish assemblies due to the higher error rates in Nanopore reads. This workflow only uses one round of polish for each of these tools. If there is a need for additional rounds of polishing, linking this workflow to another nextflow workflow is an option and instructions for doing so are covered in the linking workflows wiki page.
Donut Falls is a conservative workflow in that there are only three polishers that each are run once:
-
clair3 is a nanopore-only polisher
- Nanopore reads are mapped onto the assembly with minimap2 as part of circulocov prior to clair3
- Variants identified via clair3 are incorporated into the assembly with bcftools
-
Polypolish makes changes according to mapped reads.
- Illumina reads are mapped onto the assembly with bwa prior to polypolish
- pypolca makes changes according to mapped reads
Both polypolish and pypolca require the input of Illumina reads.
All draft genomes in their varying states of polishing are sent to busco for comparison.
Hybrid assembly does not use the polishing subworkflow.
---
Donut Falls
---
flowchart LR
A[assembled genome] --> B[circulocov for minimap2 assembly]
B --> C[clair3]
C --> D[bcftools for incorporation]
D --> E[bwa]
E --> F[polypolish]
F --> G[pypolca]
G --> H[consensus fasta]
I[illumina fastq] --> E
I --> G
