unipept: 3.1

Switch to UniProt proteomes as data source for the peptidome analysis Loading proteomes is now 3-5 times faster Calculating the unique peptidome is now 5-25 times faster Downloading sequences from the unique peptide finder is now 3-5 times faster Don't abbreviate virus names in the peptidome analysis Mark type strains and reference proteomes in the proteome selector Use better default settings in the makefile to build the database Better invalidation rules for viruses Fix a bug where the API could return an error 500 in an obscure case Fix a bug where the invalid input would result in an error 500 when converting peptides Fix a bug where the TPA would crash for non-ascii input Update ruby on rails to 4.2.