exonerate_hits.py returning reverse-complement sequences

[mossmatters]

A user identified a seemingly random bug where half of the sequences were very different. Upon closer inspection, the sequences were reverse complement.

This is a result of addressing: #27 in January 2018. Sequences recovered before the update to 1.3 should not be affected.

This should not have been fixed. When there is one single hit, HybPiper should return the exonerate results without modification. Strandedness is only needed when building a "supercontig" from multiple exonerate hits.

Users who recovered sequences with HybPiper 1.3 can fix their sequences by updating HybPiper to 1.3.1 (for example with git pull) and re-running just the exonerate step. Add the following flags to read_first.py:

reads_first.py --prefix sampleName --no-blast --no-assemble --no-distribute