Download all hg19 coding sequences from UCSC

Question

[Daniel]

How to download all human coding sequences from UCSC table browser. The resulting format that we want to send to Galaxy is "gene ID, CDS in fasta".

Answer 1

[Dan D]

on the bash command line (assuming you have a mysql client installed):

mysql --user=genome --host=genome-mysql.cse.ucsc.edu -A -D hg19 \
-N -e 'SELECT * FROM knownGeneMrna' | sed -e 's/^/>/' -e 's/\s/\n/' > myFastaFile.fa

This will take a while to run. Make sure the output is what you want by sticking a LIMIT 10 in your SQL query:

mysql --user=genome --host=genome-mysql.cse.ucsc.edu -A -D hg19 \
-N -e 'SELECT * FROM knownGeneMrna LIMIT 10' | sed -e 's/^/>/' -e 's/\s/\n/' > myFastaTestFile.fa

[Daniel]

Thanks. Any way to do it through the table browser web interface?

[Devon Ryan]

If you're willing to use the Ensembl annotation then you can just use Biomart.

[jacobsen.jeremy]

Doesn't this just download knownGeneMrna? These aren't just CDSs but include UTRs. Is there a way to get the CDS?

[Devon Ryan]

There's a "coding sequence" option.