This is a test version of Biostars. For the public version, visit https://www.biostars.org.
RefSeq assession IDs

Hi,

I am using Entrez.esearch in python to find genomic sequence of a gene (APRT gene):

from Bio import Entrez
handle=Entrez.esearch(db="nucleotide", term="APRT [Gene] AND Homo sapien [Organism] AND RefSeq[Filter]")
record=Entrez.read(handle)

In doing so, I received multiple results, including some records with accession IDs starting with NM, NC, NG, KR, CM, AY, etc. Since I am looking for genomic sequence, I think the best ones would be those starting with NG, but there are still three NG entries, including NG_008667, NG_008013 and NG_028266.

Can someone explain what are the differences between these entries and if I should filter with something more? I looked up these accession ID but I cannot tell what are the differences and which one is the best.

Thank you!

refseq gene
• 1,433 views • link
written by chaochao • chaochao

1 answer

If you want the sequence of genomic region then using EntrezDirect you can do (sequence is on minus strand so getting rev complement:

$ esearch -db gene -query "APRT [gene] AND human [orgn]" | efetch -format tabular | awk -F "\t" '{OFS="\t"}($1 == "9606"){print $12,$13,$14}' | xargs -n 3 sh -c 'efetch -db nuccore -id "$0" -seq_start "$1" -seq_stop "$2" -revcomp -format fasta'
>NC_000016.10:c88811928-88809339 Homo sapiens chromosome 16, GRCh38.p13 Primary Assembly
GGGCTGCCGCTGGCTCTTCGCACGCGGCCATGGCCGACTCCGAGCTGCAGCTGGTTGAGCAGCGGATCCG
CAGCTTCCCCGACTTCCCCACCCCAGGCGTGGTATTCAGGTGCACGCACAGGCCGCCCTCGTGGCGCCCC

If you want just the RefSeqGene entry (sequence sequence truncated for space)

$ esearch -db gene -query "APRT [gene] AND human [orgn]" | elink -db gene -target nuccore -name gene_nuccore_refseqgene | efetch -format fasta
>NG_008013.1 Homo sapiens adenine phosphoribosyltransferase (APRT), RefSeqGene on chromosome 16
GGGCCGTCGCTCACCTGTTTACACGGGCTGGGCGTGGCTGCCCACAGCCCCTGGATCTGCCGCGCAGGAT
TCGGGAAGAAGGCCCCTCGGCAGCTGCAGACTTCAGCCTGGGCTCCTGCTGTGCGGGCGAAAAGGCCCAG

If you want the RefSeq transcript sequences (sequence truncated for space)

$ esearch -db gene -query "APRT [gene] AND human [orgn]" | elink -db gene -target nuccore -name gene_nuccore_refseqrna | efetch -format fasta
>NM_001030018.2 Homo sapiens adenine phosphoribosyltransferase (APRT), transcript variant 2, mRNA
GGGCTGCCGCTGGCTCTTCGCACGCGGCCATGGCCGACTCCGAGCTGCAGCTGGTTGAGCAGCGGATCCG
CAGCTTCCCCGACTTCCCCACCCCAGGCGTGGTATTCAGGGACATCTCGCCCGTCCTGAAGGACCCCGCC

>NM_000485.3 Homo sapiens adenine phosphoribosyltransferase (APRT), transcript variant 1, mRNA
GGGCTGCCGCTGGCTCTTCGCACGCGGCCATGGCCGACTCCGAGCTGCAGCTGGTTGAGCAGCGGATCCG
CAGCTTCCCCGACTTCCCCACCCCAGGCGTGGTATTCAGGGACATCTCGCCCGTCCTGAAGGACCCCGCC
• 0 views • link
written by GenoMax GenoMax

Thank you! Do you know why the result will include CDT1 and GALNS using Entrez.esearch? Or Is there anyway I can eliminate those?

• 0 views • link
written by chaochao • chaochao

Log in to answer this question.

More posts like this