Skip to main content
Fig. 3 | Algorithms for Molecular Biology

Fig. 3

From: The paralog-to-contig assignment problem: high quality gene models from fragmented assemblies

Fig. 3

Accuracy and running time of the ExonMatchSolver on simulated data. a Dependence of the accuracy on the fragmentation level in comparison with a greedy approach. Eight paralogs, each possessing 12 exons, were simulated in two species using ALF with branch length \(n=50\) for generation of paralogs and branch length \(n=20\) for evolution of single exons. Fragmentation of exons across units was simulated 1000 times for each fragmentation level. b Dependence of the running time on paralog and exon number. Color changes of contour lines from yellow to dark blue indicate an increase in running time. Contour lines are labeled with the \(\log _{10}\) of the running time. Different numbers of paralogs (4, 6, 8, 10, 20, 40, 70, 100) and exons (1, 3, 5, 7, 10, 12, 20, 50, 70, 100) were simulated using ALF with parameters specified in the ”Methods” section and 7.7 exons per fragment on average. Running time was estimated as the mean of the user time of 20 runs with different fragmentation levels of the same simulated sequence data

Back to article page