Miso: What Does Ucsc Table Browser’S Version Of Ensembl Genes Mean ?

Question

[samsara]

I am using MISO for isoform-centric analysis of mouse rnaseq data which has later been mapped with reference genome mm10 (UCSC). I am pretty confused about the annotation they provide in their documentation page. In this link http://genes.mit.edu/burgelab/miso/docs/#human-mouse-gene-models-for-isoform-centric-analyses they have provided GFF3 annotations based on UCSC Table Browser’s version of Ensembl genes.

What does UCSC Table Browser’s version of Ensembl genes mean ?

Does it mean that the annotations are from UCSC genome with ENSEMBL geneIDs ?

Answer 1

[Jordan]

I think the general warning in the link u provided is that, make sure the chromosome names in your BAM file and the gff3 file you are using are the same.

So basically type this command:

samtools view file.bam | head | cut -f 3

If the output is something like

chr1
chr1
chr1
chr1
chr1
chr1
chr1
chr1
chr1
chr1

Then, make sure that gff3 has the same chromosome names as well. To find that out, you can just give something like this:

grep -v "#" file.gff3 | head | cut -f 1

You are good, if you get the following output:

chr1
chr1
chr1
chr1
chr1
chr1
chr1
chr1
chr1
chr1

But instead only 1 is printed, then you should be looking for a gff file which has same chr names as your BAM file. In this case UCSC gives chr names labeled as chr1 and not 1 like ensembl.

[samsara]

Did not help. I know this and this is not what i am looking for. It wish you had read page except the warning part.

Answer 2

[Emily]

UCSC display Ensembl genes. These are the genes that were originally annotated by Ensembl and imported by UCSC. All of the data in UCSC can be extracted using their Table Browser, including Ensembl genes. It would appear that MISO have done this to get their data.