Force pcr.seqs to preserve the order of the reference

[mothur-westcott]

The filter.seqs command was showing variability. It is caused by the alignment being slightly different. When the pcr.seqs command runs each processor writes to a shared output file. This sharing of the file can cause the sequences to be printed in different orders. Then when the align.seqs command runs the closest sequence in the database is the first highest result.

[mothur-westcott]

@pschloss, adding the order preservation increases the time significantly. The sorting of the file takes longer than the pcr.seqs trimming. Do you want me to look into the time increase caused by the randomization of ties in align.seqs?

[mothur-westcott]

Remove sorting and make note of variability