Polish a de-novo assemblyΒΆ

Suppose you have assembled a genome, but suspect it has some minor errors. For example, it might have been assembled from 454 or IonTorrent reads and have homopolymer length errors. The assembly can then be polished using, say, an Illumina read set (higher quality but shorter).

Method 1:

Follow the instructions in Analyse a single sample. A file consensus_masked.fa is produced with SNPs and indels patched into the reference sequences. In this file, parts of the genome with insufficient coverage to be sure of the base calling are given in lower case.

You may wish to reduce the parameter reconsensus: --cutoff, in order to favour the Illumina read set even if the evidence from it is fairly weak.

Method 2:

Follow the instructions in Variant calling for a set of samples. A file with the variants detected patched into the reference sequences is produced as part of the output.

Again, you may wish to reduce vcf-filter: --min-gq to favour using whatever evidence there is in the Illumina reads.