Large feature files need to be indexed before you can use them properly in GenomeView.

The definition of large is not strict in the sense that it depends on both the real size of the file, as well as the number of features in the file.

Recommendations:

• Only index feature files that are larger than 5 Mb or even 10 Mb when compressed with GZIP.
• Only index feature files that do contain composite features, i.e. feature consisting of multiple locations. The prime example are eukaryote genes.
• Typically the annotation of a genome does not need to be indexed if it just contains genes, mRNA, CDS and exons

Instructions:
To index a file, you need to pre-process it with tabix, much like is done with pile-up files.

Tabix can be downloaded from the tabix download page.

For BED formatted files:

sort -k1,1 -k2,2n input.bed | bgzip -c > compressed.bed.gz
tabix -p bed compressed.bed.gz


Note that indexing will not work with BED files that have a UCSC header ("track name=blah")

For GFF formatted files:

sort -T /group/tmp -k1,1 -k4,4n input.gff | bgzip -c > compressed.gff.gz
tabix -p gff compressed.gff.gz


In both cases, you will get two new files: (1) a gz file and (2) a tbi file.
Load the gz file in GenomeView.

Caveat:
The structure of genes will be lost when indexing gff files.