Hi!
We have recently started using mothur in our laboratory for analyzing antibody libraries with 454 sequencing. Mothur is an essential tool for us. I have a question. When we analyze our libraries, we get our DNA sequences with sffinfo, then we trim them by length, barcode and other properties.
The resulting fasta files are fed into another very useful software called vdjfasta. This little srcipt translates all the DNA sequences and actually finds all the hyper variable regions of the antibodies. It is very neat.
So now we have fasta files containing 60.000-100.000 short (10-30 aa) peptide sequences. We have a hard time clustering the amino acid sequences using mothur. It needs lots and lots of memory.
Since we have lots of repetitions of each clone in the dataset, it would be very useful to filter the dataset first for unique sequences. As far as I know, right now unique.seqs only works on DNA sequences. I am wondering if you or someone has modified mothur for amino acid sequences.
Csaba