Protein coding annotation from UCSC

Question

[tim.ivanov.92]

How do i download only protein-coding genes from UCSC table browser? I've chosen gencode v19 and genes and gene prediction in settings bars, but i'm getting in the result much more genes, than just protein coding

I'm attaching here screenshots with my settings

and the head of resulting dataframe

[GenoMax]

You seem to have pasted the link for image hosting site twice. Please go under "Embed codes" tab and use the full image html link and paste here.

[tim.ivanov.92]

yes, sry, fixed it in the original post - now the second picture is the right one

Answer 1

[lshepard]

One way is to choose "selected fields from primary and related tables", click the linked table and "allow selection from checked tables", selects "geneType/BioType of gene" (and anything else relevant to you), save the results and then subset/filter your file to contain rows which match protein_coding (a simple grep could do this).

I am sure there is more than one solution to the above, but this would work.

[tim.ivanov.92]

Thank you, that does it!

Answer 2

[Luis Nassar]

Hello Tim,

With your same selections on the Table Browser, if you change the output format to BED, you will see additional options to refine the output. The following page will say:

Create one BED record per:

Which by default makes an entry for the Whole Gene. You can instead designate just Exons, or only Coding Exons to exclude 3'/5' UTRs. The output will be in BED format (https://genome.ucsc.edu/FAQ/FAQformat.html#format1).

There are many answers to this question depending on exactly what you are looking for. If you have additional questions I would encourage you to look at our mailing list archives (https://groups.google.com/a/soe.ucsc.edu/forum/#!forum/genome), or write in directly to the help desk ([e-mail]).

p.s. If you are looking for a more concise gene list, you may use the UCSC Genes track for hg19, then select the knownCanonical table. This data table only has a single isoform for each gene.