Fig. 2

H. sapiens subsampled coverage QUAST results. NGA50 and misassemblies plots for a \(2 \times 51\) bp and a \(2 \times 250\) bp human dataset. The text labels indicate the offset between \(k_{rresolver}\) and \(k_{assembly}\) used for each data point. Each subplot ABySS base assembly uses optimal \(k_{assembly}\) value. As in Fig. 1, the highest \(k_{rresolver}\) is a good choice for \(2 \times 151\) bp datasets, and an offset of \(+\) 60 works well for \(2 \times 250\) bp, giving a good contiguity improvement without increasing misassemblies too much