Hi.

First thank you for any advice.

I am using external software that processes Ensmbl_GRCh38.rel80 for homo sapiens for data processing.

When I bring the data into R, the annotation that is available for H.Sapiens is rel79.

This is a discrepancy that I wish to clear up.

Should I only process data that is relevant to supported annotation libraries from bioConductor?  should I ignore these version differences (I think not) ?

Suggestions Welcome

Anthony Colombo

Alternatively, you can use the ensembldb package. Versions 75 and 79 are available through Bioconductor, but it's really simple to generate annotation packages/databases based on any Ensembl version using ensembldb and the AnnotationHub package (check the ensembldb package vignette for alternative options):

library(AnnotationHub)
library(ensembldb)
ah <- AnnotationHub()

## query AnnotationHub for available Ensembl gtf files for Ensembl release 80
query(ah, c("Homo sapiens", "release-80"))

## get the version 80 gtf:
gtf <- ah[["AH47066"]]

## generate the annotation database
DbFile <- ensDbFromGRanges(gtf, organism="Homo_sapiens", version=80, genomeVersion="GRCh38")

## we can either generate a database package using the makeEnsembldbPackage
## , or directly load the data
Edb <- EnsDb(DbFile)

## you can then use e.g. genes to get all annotations from all genes
genes(Edb)
GRanges object with 65217 ranges and 5 metadata columns:
                            seqnames                 ranges strand   |
                               <Rle>              <IRanges>  <Rle>   |
  ENSG00000000003                  X [100627109, 100639991]      -   |
  ENSG00000000005                  X [100584802, 100599885]      +   |
  ENSG00000000419                 20 [ 50934867,  50958555]      -   |
  ENSG00000000457                  1 [169849631, 169894267]      -   |
  ENSG00000000460                  1 [169662007, 169854080]      +   |
              ...                ...                    ...    ... ...
  ENSG00000281918                  1 [113079537, 113079847]      +   |
  ENSG00000281919  CHR_HSCHR5_6_CTG1 [ 33946602,  33956490]      -   |
  ENSG00000281920                  2 [ 65623272,  65628424]      +   |
  ENSG00000281921                  3 [134261776, 134261911]      +   |
  ENSG00000281922 CHR_HSCHR17_1_CTG5 [ 46784842,  46785913]      -   |
                          gene_id     gene_name  entrezid         gene_biotype
                      <character>   <character> <integer>          <character>
  ENSG00000000003 ENSG00000000003        TSPAN6      <NA>       protein_coding
  ENSG00000000005 ENSG00000000005          TNMD      <NA>       protein_coding
  ENSG00000000419 ENSG00000000419          DPM1      <NA>       protein_coding
  ENSG00000000457 ENSG00000000457         SCYL3      <NA>       protein_coding
  ENSG00000000460 ENSG00000000460      C1orf112      <NA>       protein_coding
              ...             ...           ...       ...                  ...
  ENSG00000281918 ENSG00000281918   Metazoa_SRP      <NA>             misc_RNA
  ENSG00000281919 ENSG00000281919       SLC45A2      <NA>       protein_coding
  ENSG00000281920 ENSG00000281920 RP11-418H16.1      <NA>              lincRNA
  ENSG00000281921 ENSG00000281921    AC096967.1      <NA>                miRNA
  ENSG00000281922 ENSG00000281922 RP11-1070B7.2      <NA> processed_pseudogene
                  seq_coord_system
                         <integer>
  ENSG00000000003             <NA>
  ENSG00000000005             <NA>
  ENSG00000000419             <NA>
  ENSG00000000457             <NA>
  ENSG00000000460             <NA>
              ...              ...
  ENSG00000281918             <NA>
  ENSG00000281919             <NA>
  ENSG00000281920             <NA>
  ENSG00000281921             <NA>
  ENSG00000281922             <NA>
  -------
  seqinfo: 312 sequences from GRCh38 genome
## check the vignette of the package for additional infos (e.g. filter the result, get sequences etc)

cheers, jo

A possibility is to annotate with Ensembl 80 using the biomaRt package. Take a look at this relevant recent post: Ensembl release 80 is out!