geneSymbol2Entrez.py

#! /usr/bin/env python

# ID mapping using mygene
# https://pypi.python.org/pypi/mygene
# http://nbviewer.ipython.org/gist/newgene/6771106
# http://mygene-py.readthedocs.org/en/latest/
# 08/30/14

__author__ = 'tommy'

import mygene
import fileinput
import sys


mg = mygene.MyGeneInfo()

# mapping gene symbols to Entrez gene ids and Ensemble gene ids.
# fileinput will loop through all the lines in the input specified as file names given in command-line arguments,
# or the standard input if no arguments are provided.

# build a list from an input file with one gene name in each line
def get_gene_symbols():
    gene_symbols = []
    for line in fileinput.input():
        gene_symbol = line.strip()  # assume each line contains only one gene symbol
        gene_symbols.append(gene_symbol)
    fileinput.close()
    return gene_symbols


Entrez_ids = mg.querymany(get_gene_symbols(), scopes='symbol', fields='entrezgene, ensembl.gene', species='human',
                          as_dataframe=True, verbose=False)

# set as_dataframe to True will return a pandas dataframe object, verbose=False suppress the messages like "finished".

# Entrez_ids.to_csv(sys.stdout, sep="\t")  # write the dataframe to stdout, but will not have NaNs on the screen
# if no matches were found

sys.stdout.write(Entrez_ids.to_string())  # sys.stdout.write()  expects the character buffer object

# Entrez_ids.to_csv("Entrez_ids.txt", sep="\t")    # write the pandas dataframe to csv

I didn't know that mygene existed until I came across this blog post on friday.

This was greatly helpful. Also, I created a gist. https://gist.github.com/sdhutchins/57db1bd1d979bb5408241212dfc1aec9