Automation of primer design tasks for the Buckley Lab. 

Author: Cass Johnston <[email protected]>
        Matt Burney  

We use Primer3 to design our primers. 

We use Unafold's melt.pl to check the self-hybridisation of the primers. Doesn't work so well on longer sequences so I'm ignoring amplicon self-hybridisation at the moment. Am pondering using some implementation of McCaskill's DP algorithm for partition function calculation on amplicon sequences so we can at least get probabilites of single-strandedness of bases in the amplicon.

We're currently using web interfaces to check what primers are hitting in the genome (USCSs in silico PCR). Really need to automate this.

We get sequences from Ensembl usually. This can be by gene or transcript id or by genome position. Sometimes we just have the sequence.

We have different primer design tasks, for example:
cDNA detection primers - should ideally cross an exon boundary so as to avoid amplifying genomic contamination.
genomic primers - typically for ChIP. These often need to target particular genic features, for example transcription start/end sites, splice sites, enhancer sites and so forth. 
Other sorts of primers - for example messing about with plasmids.

We often need to design many primers at the same time, for example for ChIPseq target validation. 

We'll need a basic command-line interface.

We'll also need a web interface with user-accounts that allows people to manage their primers and allows me to track the success of primers for different tasks and potentially modify primer3 parameters etc. to suit.