Multiple FASTA files and mzIdentML

[lutzfischer]

Currently there is no reference in the database which fasta-file a protein comes from. As a result xiFDR can't export the correct link for the source of a protein definition.
Currently all proteins get assigned to the first encountered database. Need to update the dataabse schema to link the protein table to the sequence file table and also xisearch to actually keep track of this during reading of the fasta files.

[lutzfischer]

need to change the db:
ALTER TABLE protein ADD COLUMN seq_id INTEGER CONSTRAINT fk_protein_sequence_file REFERENCES sequence_file(id);
so that proteins link the file in which they are defined