Indexing 5GB file

You're fortunately not the first one to store genomic information in Elasticsearch! You may be interested in reaching out to others to hear what they've done if any of these sound familiar to your own problems:

As to how to specifically deal with this, I think the answer is going to depend on what your search goal(s) is/are. 5GB of text is just going to be too much to reasonably index/search in a single document, so I think some other strategy is going to be necessary.