Update GenBLAST to take 10 best hits

[leannehaggerty]

Currently the code only takes to top ranked hit

ensembl-anno/ensembl_anno.py

Line 1521 in 926bd25

"-r",

This was because the proteins were likely to be distant from the relative and so it was likely only worth using the top hit, however this removes the potential to find paralogues

We should update this to allow the top 10 ranked hits and then introduce filtering of "bad" alignments in the code that converts the raw genblast output (this is currently written to work with only 1 hit)

ensembl-anno/ensembl_anno.py

Line 1586 in 926bd25

def convert_gff_to_gtf(gff_file):