Appending sequence decoys and virus sequences to reference genome (GRCh38)

Question

[newbio17]

Hello,

While I was experimenting with Genomic Data Commons (GDC) pipelines, I noticed that the reference genome that they have available for download is based on GRCh38 patch from 2013.

Would it be okay to append sequence decoys (hs38d1) and virus sequences to the latest GRCh38 in attempt to create an "updated" version of the reference genome used in GDC pipelines?

The sequence decoys and virus sequences shown on this page.

Also, NCBI says that hs38d1 from 2014 is the latest version of the decoys. Should I be looking for decoy sequences from somewhere else?

Thank you.

[kmzhou4]

This is a good question. I am trying to figure out whether to use the latest GRCh38.p14 (release 2022), or the hs38d1 (decoy) for my analysis. To my understanding, decopy masked the telomeric sequences, and removed the alternative locus, added the EBV virus sequence. But trying to understand what are the benefits for using each version.

Answer 1

[WouterDeCoster]

GRCh38 will always have the same sequence, regardless of the version/patch. The primary sequence hasn't changed - but the annotation might. So the GRCh38 patch from 2013 with decoys is still valid.

[kmzhou4]

To my understanding, the release will update the patch, fix, etc. The primary sequence are preserved so that within major version 38, the coordinate are the same. It is like using the same map. The patches are like more zoomed in version of insert maps with some modification or updates to the major. Is there a corresponding decoy updated version for GRCh38? the current GRCh38decoy https://www.ncbi.nlm.nih.gov/assembly/GCA_000786075.2/ is labeled as partial (only 5.8 MBases). Is there any overlap or conflict between the patches from GRCh38 and the GRCh38decoy?