Seperating sequences with single base pair mutations from a cDNA library - (Oct/07/2013 )

I have some cells that are heterozygous for about 400 sequences that I want to isolate, and each allele differs by only one base pair.  RNAseq suggests they're all transcribed, so I've generated a cDNA library from these cells.  

What is the best way to isolate only the mutant sequences? 

So far, the only thing I've come up with is to PCR out each transcript of interest from my cDNA library, run on gel and cut out band, clone into vector, transform into bacteria, and pick colonies and sequence until I find my mutant allele.  

If you have access to one, a dHPLC will separate out these sorts of things from the PCR - it would save you all the cloning etc.