Tuxedo bug report: H. saltator simulated data

I randomly selected a 150kb region from Harpegnathos saltator to create a demo data set for instruction. I extracted sequences for the transcripts encoded in the region and used wgsim to simulate Illumina reads for these transcripts. The idea was to use the Tuxedo workflow (Bowtie + Tophat + Cufflinks) to map the reads back to the genome, reconstruct transcripts, and do differential expression analysis. Unfortunately the latest version of Cufflinks (as of Feb 3 2015) fails for this data set, crashing with a segfault and without any helpful error message.