inpIDMapper: Convenience functions for mapping IDs through an appropriate...

In Bioconductor/AnnotationDbi: Manipulation of SQLite-based annotations in Bioconductor

Convenience functions for mapping IDs through an appropriate set of annotation packages

Description

These are a set of convenience functions that attempt to take a list of IDs along with some addional information about what those IDs are, what type of ID you would like them to be, as well as some information about what species they are from and what species you would like them to be from and then attempts to the simplest possible conversion using the organism and possible inparanoid annotation packages. By default, this function will drop ambiguous matches from the results. Please see the details section for more information about the parameters that can affect this. If a more complex treatment of how to handle multiple matches is required, then it is likely that a less convenient approach will be necessary.

Usage

  inpIDMapper(ids, srcSpecies, destSpecies, srcIDType="UNIPROT",
  destIDType="EG", keepMultGeneMatches=FALSE, keepMultProtMatches=FALSE,
  keepMultDestIDMatches = TRUE)

  intraIDMapper(ids, species, srcIDType="UNIPROT", destIDType="EG",
  keepMultGeneMatches=FALSE)

  idConverter(ids, srcSpecies, destSpecies, srcIDType="UNIPROT",
  destIDType="EG", keepMultGeneMatches=FALSE, keepMultProtMatches=FALSE,
  keepMultDestIDMatches = TRUE)

Arguments

ids	a list or vector of original IDs to match
srcSpecies	The original source species in in paranoid format. In other words, the 3 letters of the genus followed by 2 letters of the species in all caps. Ie. 'HOMSA' is for Homo sapiens etc.
destSpecies	the destination species in inparanoid format
species	the species involved
srcIDType	The source ID type written exactly as it would be used in a mapping name for an eg package. So for example, 'UNIPROT' is how the uniprot mappings are always written, so we keep that convention here.
destIDType	the destination ID, written the same way as you would write the srcIDType. By default this is set to "EG" for entrez gene IDs
keepMultGeneMatches	Do you want to try and keep the 1st ID in those ambiguous cases where more than one protein is suggested? (You probably want to filter them out - hence the default is FALSE)
keepMultProtMatches	Do you want to try and keep the 1st ID in those ambiguous cases where more than one protein is suggested? (default = FALSE)
keepMultDestIDMatches	If you have mapped to a destination ID OTHER than an entrez gene ID, then it is possible that there may be multiple answers. Do you want to keep all of these or only return the 1st one? (default = TRUE)

Details

inpIDMapper - This is a convenience function for getting an ID from one species mapped to an ID type of your choice from another organism of your choice. The only mappings used to do this are the mappings that are scored as 100 according to the inparanoid algorithm. This function automatically tries to join IDs by using FIVE different mappings in the sequence that follows:

1) initial IDs -> src organism Entrez Gene IDs 2) src organism Entrez Gene IDs -> sre organism Inparanoid ID 3) src organism Inparanoid ID -> dest organism Inparanoid ID 4) dest organism Inparanoid ID -> dest organism Entrez Gene ID 5) dest organism Entrez Gene ID -> final destination organism ID

You can simplify this mapping as a series of steps like this:

srcIDs —> srcEGs —> srcInp —> destInp —> destEGs —> destIDs (1) (2) (3) (4) (5)

There are two steps in this process where multiple mappings can really interfere with getting a clear answer. It's no coincidence that these are also adjacent to the two places where we have to tie the identity to a single gene for each organism. When this happens, any ambiguity is confounding. Preceding step \#2, it is critical that we only have ONE entrez gene ID per initial ID, and the parameter keepMultGeneMatches can be used to toggle whether to drop any ambiguous matches (the default) or to keep the 1st one in the hope of getting an additional hit. A similar thing is done preceding step \#4, where we have to be sure that the protein IDs we are getting back have all mapped to only one gene. We allow you to use the keepMultProtMatches parameter to make the same kind of decision as in step \#2, again, the default is to drop anything that is ambiguous.

intraIDMapper - This is a convenience function to map within an organism and so it has a much simpler job to do. It will either map through one mapping or two depending whether the source ID or destination ID is a central ID for the relevant organism package. If the answer is neither, then two mappings will be needed.

idConverter - This is mostly for convenient usage of these functions by developers. It is just a wrapper function that can pass along all the parameters to the appropriate function (intraIDMapper or inpIDMapper). It decides which function to call based on the source and destination organism. The disadvantage to using this function all the time is just that more of the parameters have to be filled out each time.

Value

a list where the names of each element are the elements of the original list you passed in, and the values are the matching results. Elements that do not have a match are not returned. If you want things to align you can do some bookeeping.

Author(s)

Marc Carlson

Examples

## Not run: 
  ## This has to be in a dontrun block because otherwise I would have to
  ## expand the DEPENDS field for AnnotationDbi
  ## library("org.Hs.eg.db")
  ## library("org.Mm.eg.db")
  ## library("org.Sc.eg.db")
  ## library("hom.Hs.inp.db")
  ## library("hom.Mm.inp.db")
  ## library("hom.Sc.inp.db")

  ##Some IDs just for the example
  library("org.Hs.eg.db")
  ids = as.list(org.Hs.egUNIPROT)[10000:10500] ##get some ragged IDs
  ## Get entrez gene IDs (default) for uniprot IDs mapping from human to mouse.
  MouseEGs = inpIDMapper(ids, "HOMSA", "MUSMU")
  ##Get yeast uniprot IDs in exchange for uniprot IDs from human
  YeastUPs = inpIDMapper(ids, "HOMSA", "SACCE", destIDType="UNIPROT")
  ##Get yeast uniprot IDs but only return one ID per initial ID
  YeastUPSingles = inpIDMapper(ids, "HOMSA", "SACCE", destIDType="UNIPROT", keepMultDestIDMatches = FALSE)

  ##Test out the intraIDMapper function:
  HumanEGs = intraIDMapper(ids, species="HOMSA", srcIDType="UNIPROT",
  destIDType="EG")
  HumanPATHs = intraIDMapper(ids, species="HOMSA", srcIDType="UNIPROT",
  destIDType="PATH")

  ##Test out the wrapper function
  MousePATHs = idConverter(MouseEGs, srcSpecies="MUSMU", destSpecies="MUSMU",
  srcIDType="EG", destIDType="PATH")
  ##Convert from Yeast uniprot IDs to Human entrez gene IDs.
  HumanEGs = idConverter(YeastUPSingles, "SACCE", "HOMSA")


## End(Not run)

Bioconductor/AnnotationDbi documentation built on Nov. 2, 2024, 6:27 a.m.