Comparison of E. coli MDA assemblies to the expected (ideal) genomic span. Newbler assembly with all reads (ALL) fails on all samples during consensus resolution (not shown). SPAdes is able to complete an assembly using all reads with few missing reference bases in the assembly but the output also generates excessive redundant and spurious contigs. Pre-processing reduces sequences allowing the completion of OLC (Newbler) assembly and yields genomic spans by both Newbler and SPAdes which are most similar to the ideal (reference) assembly. The Newbler assembly from coverage-reduced data is also closer to the reference span and verifies (consistent with) the assembly produced by the SPAdes following preprocessing alone.