Does AnnotationHub replaces AnnotationDbi for ID conversion (e.g. entrez, refseq, symbols)?
1
1
Entering edit mode
dalloliogm ▴ 40
@dalloliogm-7141
Last seen 5 months ago
European Union

Hello,

I am exploring the new AnnotationHub package and I find it fantastic. Thank you for all the effort, and also for the nice tutorials on coursera and youtube.

I do not understand how much AnnotationHub will replace existing packages such as AnnotationDbi or the organism packages. For example I wonder if there is a way to convert an Entrez Id to convert gene Entrez to symbols and other IDs using AnnotationHub, or if I still have to rely on AnnotationDbi and the organism packages. In general I have a bit of confusion regarding which is the correct way to convert gene id from one type to another in bioconductor, without using biomaRt. Can anyone please help me clarify it?

Thanks

conversion symbol entrez • 1.0k views
2
Entering edit mode
@james-w-macdonald-5106
Last seen 2 minutes ago
United States

The AnnotationHub is just an easier way to supply annotation data to end users. It doesn't change how you interact with the data that you download - it just changes how you search for and retrieve the data. For example, if there were an org.Hs.eg.db on the AnnotationHub (I don't believe there is, btw), then you would still need AnnotationDbi to query it. As an example, for an OrgDb package that is on the hub:

> query(hub, c("OrgDb"))
AnnotationHub with 1145 records
# snapshotDate(): 2015-08-26
# $dataprovider: NCBI #$species: 'Nostoc azollae'_0708, Acaryochloris marina_MBIC11017, Acetobact...
# \$rdataclass: OrgDb
# additional mcols(): taxonomyid, genome, description, tags, sourceurl,
#   sourcetype
# retrieve records with, e.g., 'object[["AH12818"]]'

title
AH12818 | org.Pseudomonas_mendocina_NK-01.eg.sqlite
AH12819 | org.Streptomyces_coelicolor_A3(2).eg.sqlite
AH12820 | org.Cricetulus_griseus.eg.sqlite
AH12821 | org.Streptomyces_cattleya_NRRL_8057_=_DSM_46488.eg.sqlite
AH12822 | org.Cavia_porcellus.eg.sqlite
...       ...
AH13958 | org.Ochotona_princeps.eg.sqlite
AH13959 | org.Aeromonas_veronii_B565.eg.sqlite
AH13960 | org.Oryctolagus_cuniculus.eg.sqlite
AH13961 | org.Tetraodon_nigroviridis.eg.sqlite
> z <- hub[["AH12818"]]
> z
OrgDb object:
| DBSCHEMAVERSION: 2.1
| DBSCHEMA: NOSCHEMA_DB
| ORGANISM: Pseudomonas mendocina_NK-01
| SPECIES: Pseudomonas mendocina_NK-01
| CENTRALID: GID
| TAXID: 1001585
| Db type: OrgDb
| Supporting package: AnnotationDbi

GID   SYMBOL
1 10454784 MDS_0002
2 10454785 MDS_0003
3 10454786 MDS_0004
4 10454787 MDS_0005
5 10454788 MDS_0006
6 10454789 MDS_0007

Long term I think everything is going to end up in AnnotationHub, rather than being downloaded and installed. But for now there is still a set of packages that are installed, and another set that are on the AnnotationHub. But you don't interact with the data from either source differently - you just get the data in a different way.

0
Entering edit mode

Thank you, this is exactly what I wanted to know. I was surprised to see that the human OrgDb is not available, but I guess it will be introduced in the future.

0
Entering edit mode

Just to clarify and avoid any confusion: anything you get from AnnotationHub is also downloaded and installed. With AnnotationHub things end up in ~/.AnnotationHub/ (the AnnotationHub "cache"). Traditional annotation packages end up in your package library. Personally I'm still not totally convinced that this split is a good idea but maybe it's just me...

H.