Thanks for your detailed reply.
I had been thinking about clustering the contigs today. I have used blastclust before also I was having a look at CLOBB (
http://www.biomedcentral.com/1471-2105/3/31/abstract) as it seems to be similar to blastclust but optimised for EST clustering. Also I found a program called wcd (
http://bioinformatics.oxfordjournals...ull/24/13/1542) that is also designed EST data.
I think I'll give the clustering a go. If I have lots of problems then I might look into using the pipeline you suggested.