Downloading hg38 centromere and telomere positions from UCSC table browser

Question

[graeme.thorn]

I'm converting an R code (from here: https://github.com/cancer-genomics/delfi_scripts) from hg19 to hg38 assembly, and it relies on automatically downloading telomeric and centromeric regions from the UCSC table browser:

genome <- "hg19"
mySession <- browserSession()
genome(mySession) <- genome
gaps <- getTable(ucscTableQuery(mySession, track="gap"))

in order that the resulting fragment calculations don't cover (some of) the less mappable regions of the genome. However, the corresponding code for hg38:

genome <- "hg38"
mySession <- browserSession()
genome(mySession) <- genome
gaps <- getTable(ucscTableQuery(mySession, track="gap"))

does not return the centromere positions from the UCSC table (it has all the telomeric ranges). I have tested the online table browser and that does not return hg38 centromere positions either. Is there another source for these?

[GenoMax]

While UCSC support stops by here once in a while you should probably report this directly to them (genome at soe.ucsc.edu) and then provide an update here.

[graeme.thorn]

I will provide feedback to them about this. However, I'm just looking for a table of positions in hg38 that I can bolt on to the existing removed regions to ease the workflow.

EDIT: it does look like this is a frequent question to them, see for instance here: https://groups.google.com/a/soe.ucsc.edu/forum/#!msg/genome/SaR2y4UNrWg/XsGdMI3AazgJ

The answer to this query doesn't really help, though.

[Simo]

Hi, still any solution?

I'm looking for the centromeric regions in release GRCh38.

I would need something like: chr start end

I don't understand where I can find it, I looked everywhere and still there is no direct information.

Thanks

[ATpoint]

See my answer.

Answer 1

[ATpoint]

For centromers, the table browser is the answer.

Select BED as output format, then get output, then whole gene, then get BED, done.

[Simo]

Hi, thank you for the answer.

Eventually I got what I needed with few adjustments.

Here my solution to obtain centromeric coordinates for hg38:

• Go to the Table Browser: http://genome.ucsc.edu/cgi-bin/hgTables
• Choose the Mapping and Sequencing group
• Select the "Chromosome Band (Ideogram)" track
• Select filter, and enter "cen" in the gieStain field
• Press "submit" and then "get output"

Each chromosome will have two entries which overlap.

They can be simply merged into a single entry.

I hope this could be helpful for others.

[David Quigley]

As of 2021 the filter you want for this approach is "acen" rather than "cen", as noted by Simo above.