During the process of curation of RNA interference (RNAi) data, WormBase routinely maps the targets of any given RNAi experiment to the genome based on information present in the paper that describes the experiment. Recently WormBase has refined this process and addressed inconsistencies in target determination.  Previously, we were not filtering out the highly fragmented hits that occurred. That is, when many very short alignments occurred close together on the genome our mapping script was concatenating these splits, much like it would do when it skips over introns. These hits caused errant primary and secondary targets to be displayed. Most targets for RNAi experiments remain unchanged, but errant hits have been removed from WormBase.
The criteria for primary and secondary target determination (these descriptions are also on the individual RNAi report pages) are as follows:
Primary targets: These are targets that have sequence identity to the RNAi probe of at least 95%, over a stretch of at least 100 nucleotides, identified using a
combination of BLAST and BLAT algorithms.  These are usually the intended target genes of an RNAi experiment.
Secondary targets: These are targets that have between 80 and 94.99% sequence identity over a stretch of at least 200 nucleotides to the RNAi probe. Targets (and overlapping genes) that satisfy these criteria may or may not be susceptible to a RNAi effect with the given probe and represent secondary (unintended) genomic targets of an RNAi experiment.