I'm curious how DWGSIM defines incorrectly mapped reads. Most of the time I use DWGSIM, the number of incorrectly mapped reads, even the highest quality reads, are mappend incorrectly 20-60% of the time.

For example, in one simulated dataset, of 14,188 reads, 6,724 of them are mapped correctly (mc), and 7,464 mapped incorrectly (mi). Does this mean that the mapped locus is different from the locus that the simulated read originated from? Am I thinking of this incorrectly? How do I trust a mapper that maps over 50% of reads incorrectly?