Fig. 1

H. sapiens parameter sweep QUAST results. NGA50 and misassembly scaffold metrics with and without RResolver. High-quality assemblies lean towards top left corner, with high contiguity and low misassemblies. The text labels indicate the offset between \(k_{rresolver}\) and \(k_{assembly}\) used for each data point. Some text labels (for smaller triangles) and overlapping data points are omitted to reduce crowdedness in the plot, while keeping the trends. All RResolver data points have higher NGA50 than the corresponding baseline assembly, and some have fewer misassemblies. For \(2\) bp datasets, picking the highest \(k_{rresolver}\) increases NGA50 the most while keeping misassembly increase moderate. For \(2\) datasets, picking the highest \(k_{rresolver}\) is not necessarily optimal as it leads to increased misassemblies, and a \(k_{rresolver} = k_{assembly} + 60\) is a good empirical choice for balancing NGA50 increase and minimizing misassemblies