Skip to main content
Figure 8 | BMC Bioinformatics

Figure 8

From: Alternative splicing detection workflow needs a careful combination of sample prep and bioinformatics analysis

Figure 8

Statistical detection of spliced variants. A) True positive transcripts detected as differentially expressed between C1-C5 and T1-T5 groups as function of the spike-ins (20, 40, 80 million reads) and of the LSP (NU, TS) out of 27 spiked-in transcripts. B) True positive synthetic transcripts detected as differentially expressed between C1-C5 and T1-T5 out of the 58 differentially expressed synthetic transcripts with known fold change between T1-T5 and C1-C5. Only the 80 millions spike-in was considered since the synthetic spike-ins are identical over all the datasets. C) False positive transcripts detected in the dataset 80 NU dataset with spike-in derived from 80 million reads, depending on the BA used in the analysis. Only the 80 millions spike-in is considered since the false positive are nearly in the same amount for 20, 40 and 80 millions spikes-in. D) False positive transcripts detected in the TS dataset with spike-in derived from 80 million reads depending on the BA used in the analysis. Only the 80 millions spike-in was considered since the false positive are nearly in the same amount for 20, 40 and 80 millions spikes-in. The numbers on the top of each bar indicates the % of detectable spikes-in.

Back to article page