This would be good for a case where a virus sample has (too) high coverage - this means some parameters don't work well in the assembly (i.e. they are optimised for e.g. X100 not X10,000). (At some point, increased coverage just contributes erroneous k-mers rather than useful information).
- Use seqtk to subsample the fastq file before assembly.
- The seqtk command required is "sample"
- "seqtk sample" is in the Galaxy Toolshed but need to see if it works.
This would be good for a case where a virus sample has (too) high coverage - this means some parameters don't work well in the assembly (i.e. they are optimised for e.g. X100 not X10,000). (At some point, increased coverage just contributes erroneous k-mers rather than useful information).