Table 2.

Consensus quality of the bounded read placement algorithm.

Species	# consensus differences	True positives	False positives
E. coli O157:H7	14	14	0
S. enterica	5	5	0
B. mallei	47	44	3
I. multifiliis	N/A	N/A	N/A
E. coli K12	N/A	N/A	N/A
C. amycolatum	N/A	N/A	N/A
Total	66	63	3

Performance of the same three algorithms described in Table 1. Number of consensus differences: The total number of bases in consensus that are different between the bounded and control assemblies versus reference. True positive: Number of consensus base changes that are supported by the reference. False positive: The number of consensus base changes that differ from the reference. The finishing reads used for E. coli K12 did not come from the same strain as the reference. We cannot validate whether a consensus discrepancy between an assembly and the reference is due to assembly error or to strain-level differences. Consensus quality could not be measured on the two genomes that lack a reference.