Table 1.

RNASEG results on simulated data

		a	F-measure				Recovered sequencing data [%]
	u		1	10	100	1,000	1	10	100	1,000
t = 1 (170 h)		1	0.93	0.93	0.93	0.93	100.00	100.00	99.95	97.92
		10	-	0.93	0.92	0.93	-	100.00	99.96	98.10
		100	-	-	0.86	0.86	-	-	99.91	99.12
		1,000	-	-	-	0.67	-	-	-	97.73
t = 100 (156 h)		1	0.93	0.93	0.93	0.92	100.00	100.00	99.96	98.63
		10	-	0.93	0.92	0.92	-	100.00	99.96	98.10
		100	-	-	0.86	0.86	-	-	99.91	99.12
		1,000	-	-	-	0.67	-	-		97.73
t = 1,000 (110 h)		1	0.87	0.86	0.83	0.81	100.00	99.98	99.83	97.76
		10	-	0.86	0.83	0.81	-	99.98	99.83	97.76
		100	-	-	0.82	0.80	-	-	99.81	97.76
		1,000	-	-	-	0.67	-	-	-	97.75

F-measure and fraction of recovered sequencing data from the secondary library (in %) for simulated data. RNASEG was applied to the forward strand of region 684,676 to 987,046 of the H. pylori genome. We set $\widehat{\mathit{\text{k}}}$ = 20,000, $\widehat{\mathit{\text{cp}}}$ = 1,000, w = 100, and varying values for parameters t, a and u. We define: True positives (TP) are genes that are part of a transcript segment, true negatives (TN) are intergenic regions that are not part of a transcript segment, false positives (FP) are intergenic regions that are part of a transcript segment and false negatives (FN) are genes that are not part of a transcript segment. ‘-’ indicates parameter combinations that have not been tested because they are not sensible. Numbers in brackets below t correspond to the average runtime in CPU hours for all simulations with this value of t.