Create a custom exome .genome for IGV

Question

[jgarces]

Hi everyone,

I'm trying to create a custom file.genome for IGV, but I can't and don't know what's the error... According to IGV instructions (https://software.broadinstitute.org/software/igv/LoadGenome), I'm following these steps:

1. I've downloaded whole genome fasta and have extracted only those regions coding for genes (Is There A Way To Download The Human Exome?).
2. In IGV, I select the create option pointing out this modified fasta, a text file with cytobands and a gtf downloaded from Ensembl.
3. Generated .genome is (extremelly?) heavy (~40Mb) and IGV crashes trying to opening it (CPU is totally overhelmed).

Any idea, please?? Thanks in advance!!

[GenoMax]

This may not be the ideal way of doing this. You probably scores of "chromosomes" (which would be the regions that you are looking at) and that may be overwhelming the display in IGV which by default starts with entire set of chromosomes. You may want to use the default genomes and then use a BED interval file with regions you are interested in to navigate.

Edit: Looks like @h.mon basically wrote more or less the same answer at the time I posted this.

Answer 1

[h.mon]

I don't know if what you are doing is correct or not, but I believe its is incorrect because I guess your cytobands and gtf from ENSEMBL are based on the full genome assembly, and the reference exome you are trying to create will have thousands of "chromosomes" with no correspondence to the cytoband and gtf.

What I think you want:

Use the full reference genome (hg18, hg19, etc), download the bed with exome coordinates, and on "tracks" "file <- open file" open the exome bed, cytoband file (has to be in a formt that IGV understands) and gtf from ensembl. All files - exome bed, cytoband file and gtf - need to be from the same genome version as the reference of your choice.

[jgarces]

Thanks @h.mon but there's a thing I don't understand... you need the fasta for create the .genome (bed it's not enough), where I should introduce it?

[h.mon]

There is a drop-down menu to the upper-left of IGV interface, you can choose several genomes there.

[jgarces]

Thanks again @h.mon, maybe I'm missunderstanding you but this option only changes the annotation of genes but introns remain... I want keep only exons (for this reason I asked about a new fasta).

[h.mon]

Your problem is keeping all coordinates in sync: the exome bed, the cytoband txt (I hope it is bed) and the annotation gtf. If you convert the whole genome fasta to an exome fasta, the two other files (cytoband and gtf) will be useless, as their coordinates will have no meaning on this new reference. If you load a human reference and a bed with exome coordinates, you will see exactly where the exons are. The introns will be there, but the exons will be clearly delineated.

But this begs the question: what exactly is your purpose, your goal? What do you expect to see?

[GenoMax]

jgarces : You create a custom genome in IGV by doing "Genome" --> "Create a .genome file" --> Provide identifiers for the "genome" you want to make --> select your fasta file.

As I said before in my comment, if you have hundreds of fasta entries in your file this is going to become messy.

[vmicrobio]

so in order to do high throughput there is no ways to do it using command lines?