Skip to main content
Fig. 2 | BMC Bioinformatics

Fig. 2

From: De novo Nanopore read quality improvement using deep learning

Fig. 2

Density scatter plot showing average read quality improvement by MiniScrub versus raw reads. The X-axis shows read percent identity to the reference while the Y-axis shows read length. Raw reads are in blue while scrubbed reads are in red. The darkness of the color indicates increased “density” – more reads fall into a darker region of the graph than the lighter areas. MiniScrub scrubs out most of the low-quality segments in low quality reads while leaving high quality reads intact, increasing average read percent identity by over 3%, from 83.1 to 86.2%. Average read length decreased from 2673 bases to 1594 bases due to splitting reads where low-quality segments were removed. Reads > 25kbp have low density, and are not shown in order to keep the substantive portion of the graph relatively large

Back to article page