Allele finding and primer bias

[krdav]

The problem is that it is common practice to use primers (especially in the FR1 region) that are deliberately changing a couple of nucleotides. This might be for several reasons e.g. a) to only use a sparse set of generate primers, b) to introduce a restriction site for future subcloning etc.

The situation could look like this:

The introduction of this change will look as if it was caused by SHM and will be found in the majority (like 99%) of all the reads. Therefore it could easily be confused to be a new allele. Adding to the problem; when running data from other labs it might not be clear switch primers they have used. Additions to the code should be added to deal with this problem:

1. Switch to disable allele finding on the first (and/or last) X nucleotides.
2. Option to revert a primer introduced change (otherwise detected as a new allele) back to its germline identity.
3. Primer finding and warning by searching for synonymous mutations in the first and last 30 nt. of a read. If non are observed it is probably because a primer have reverted these.

[psathyrella]

dear self: add option to override automatic 5' and 3' allelefinder exclusion zones.