Map between gene identifiers in GO

Currently supported identifiers

Identifier	Setting name	Examples
BioGRID	Biogrid	116400
Ensembl gene	EnsemblGene	ENSG00000114209
Ensembl protein	EnsemblProtein	ENSP00000376506
Entrez gene ID	Entrez	11235
gene symbol	Symbol	PDCD10, CCM3
HGNC	HGNC	8761
Protein name	Name	Programmed cell death protein 10
Refseq mRNA	RefseqMRNA	NM_007217
Refseq protein	RefseqProtein	NP_009148
UniProt Accession	Accession	Q9BUL8
UniProt ID	ID	PDC10_HUMAN

Installation for CLI (optional)

> go get -v github.com/knightjdr/genemap
> cd $GOPATH/src/github.com/knightjdr/genemap
> go install ./...

Usage

Generate file for mapping.

The first step in mapping is to download reviewed UniProt entries and approved HGNC entries and create a JSON file containing supported conversion identifiers. Currently this only supports human genes. By default the file will be created in the current directory but a target directory can be passed as a flag/option. The created mapping file will have the creation date in the name formatted like genemap-YYYY-MM-DD.json. This allows you to have your own database for mapping and update it as you see fit.

Import

import "github.com/knightjdr/genemap"

func main() {
  genemap.Generate("/path/to/output/folder/")
}

CLI

The map generation step can be run from the command line

> genemapcreate --folder="/path/to/output/folder/"

Convert identifiers

You can use the mapping file created in the previous step to convert between supported gene identifiers. The Mapper struct is used for setting mapping parameters and will contain the results of the mapping.

Import

import "github.com/knightjdr/genemap"

func main() {
  mapper := genemap.CreateMapper()
  mapper.Load("/path/to/map/genemap.json")
  mapper.FromType = "Symbol"
  mapper.ToType = "Accession"

  err := mapper.Convert(ids)
  if err != nil {
    log.Fatal("Error converting IDs")
  }

  fmt.Println(mapper.Converted)
}

The mapper struct looks like so:

type Mapper struct {
	Converted          map[string]string
	FromType           string
	PossibleConversions map[string][]string
	ToType             string
	Unconverted        []string
}

Converted will have a one-to-one mapping of identifiers between the requested types and Unconverted will contain a list of identifiers that could not be converted. PossibleConversions will contain all possible mappings that the supplied identifier could be converted to. For example, if you request a mapping to UniProt identifiers, Converted will have a mapping to a reviewed UniProt entry while PossibleConversions will have all possible consistent UniProt entries, whether reviewed or not. When converting to UniProt accession or ID, the Converted map will always map to the reviewed entry. Similarly, when converting to gene Symbol, the entry will always map to the official gene symbol. When mapping to other conversion types, the Converted map will simply contain one of the possible identifiers it can be converted to if there are multiple.

CLI

ID conversion can be done from the command line. It requires a mapping file (.json format) and a text file with IDs to convert (one entry per line). It will output a text file with the converted identifiers. The output file name will be called conversion.txt if one is not supplied.

> genemapconvert --fromType="Symbol" --toType="Accession" --idFile="ids.txt" --mapFile="genemap.json" --outFile="my-out-file.txt"