Skip to main content
Fig. 2 | BMC Genomics

Fig. 2

From: XCAVATOR: accurate detection and genotyping of copy number variants from second and third generation whole-genome sequencing experiments

Fig. 2

SLM and FastCall performance. SLM and FastCall performance. Panels a-e compare the performance of SLM and CBS algorithms on the analysis of synthetic chromosomes. Panel a reports the area under the ROC curve of SLM and CBS for different sequencing coverages, while panel b summarize the performance of SLM and CBS in the detection of the correct breakpoint position. On the x axis is reported the distance between the predicted and the correct position. On the y axis is reported the fraction of breakpoints predicted at a given distance from the correct position. Panels c-e show TPR (c for deletions, d for duplications) and # FP e for SLM and CBS as a function of number of windows. # FP of panel e is calculated as the average number of false positive events detected in all the synthetic chromosome we simulated. Panels f-k report the resolution of SLM+FastCall algorithms for different sequencing coverages (f-h) and different window size (i-k). Resolution results are reported for deletions (f and i), duplications (g and j) and false positive events (h and k). # FP in panels h and k is estimated as the total number of false positive events we can expect from the analysis of an entire human genome

Back to article page