Polish with Reads

The Polish with Reads tool facilitates the process of refining a set of sequences with high-quality reads. This enables the creation of hybrid assemblies by first creating an assembly from long, error prone reads, and subsequently using high-quality Illumina reads to polish the contigs.

Before polishing, high-quality reads should be stripped of adapters and lower quality bases. This can be done using Trim Reads (http://resources.qiagenbioinformatics.com/manuals/clcgenomicsworkbench/current/index.php?manual=Trim_Reads.html).

The tool uses Racon [Vaser et al., 2017] with additional improvements inspired by Minipolish [Wick and Holt, 2019]. Racon uses a divide-and-conquer approach for rapid consensus calling. The partial order alignment (POA) of the reads against the target sequences occurs in non-overlapping windows on the target sequences. This approach has the consequence that Racon may not always use a globally optimal alignment of reads for consensus calling.

Polishing is conducted in two steps. Following each step, a set of corrections, inspired by minipolish [Wick and Holt, 2019], are carried out to improve contig quality:

To run the Polish with Reads tool, go to:

        Toolbox | Long Read Support (Image longreads_folder_closed_16_h_p) | Polish with Reads (Image polish_w_reads_16_h_p)

Select a sequence list containing contigs or long reads.

In the next dialog, set the polishing parameters (figure 6.1):

Image longreadpolish_step1
Figure 6.1: Polish with Reads parameters



Subsections