BSgenome.Hsapiens.UCSC.hg38 vs BSgenome.Hsapiens.NCBI.GRCh38

Question

[ZheFrench]

I am wondering what is the difference between this two reference package in R ? hg38 and GRCh38 should be same sequence assembly, right ? So what the point of creating this two reference package to access raw sequence.

[finswimmer]

Maybe the naming convention for the chromosomes?

UCSC usually prefix the chromosome with chr, NCBI doesn't.

From the examples in the manuals:

• BSgenome.Hsapiens.NCBI.GRCh38.pdf genome[["1"]]

• BSgenome.Hsapiens.UCSC.hg38.pdf genome$chr1 # same as genome[["chr1"]]

fin swimmer

[zx8754]

It is the same question as "the difference between UCSC and NCBI genomes". Why don't you load them and compare?

Answer 1

[igor]

There are actually many different versions of hg38/GRCh38. This post by lh3 describes the human reference genome landscape very nicely: http://lh3.github.io/2017/11/13/which-human-reference-genome-to-use

Specific discrepancies:

• Inclusion of ALT contigs.
• Padding ALT contigs with long “N”s.
• Inclusion of multi-placed sequences.
• Not using the rCRS mitochondrial sequence.
• Converting semi-ambiguous IUB codes to “N”.
• Using accession numbers instead of chromosome names.
• Not including unplaced and unlocalized contigs.