Figure 1.
Distribution of sequencing errors.
(A) shows the sequencing errors occur randomly across one read and B) shows the error rate is almost the same among six human datasets with different read lengths.
Figure 2.
Assessing the quality of de novo assemblies on human datasets with different lengths using four different measures: (A) the percentage of full-length reconstructed reference transcripts, (B) false positive rate, (C) nucleotide sensitivity, and (D) nucleotide specificity.
Figure 3.
Four assessment metrics of assemblies on S.cerevisiae datasets with different lengths.
Figure 4.
Assessing the quality of de novo assemblies on mouse datasets with different lengths.
Table 1.
Comparison of de novo assemblies on S.cerevisiae datasets with different sequencing error rates (the read length is 75 bp).
Figure 5.
Analyses of five human datasets with different numbers of spliced isoforms.
(A) Boxplot of exon number of each gene. (B) Boxplot of spliced isoform length.
Table 2.
Statistics of five human datasets with different numbers of spliced isoforms.
Figure 6.
Comparison of de novo assemblies on five human datasets with different numbers of spliced isoforms.