How do I get the GFF output file from FIMO onto UCSC Genome Browser?

[eele...@gmail.com]

I ran FIMO and got a GFF file.  But when I try to load the GFF into UCSC Genome Browser, I get an error message. 

If I look at the GFF output as a text file, it looks generally OK.  But as you can see from column #1 below, there are no coordinates for the features, just 'danRer11_dna'.  But the coordinates were there in the input data:

GFF OUTPUT:

##gff-version 3
danRer11_dna    fimo    nucleotide_motif    9598    9616    53.7    -    .    Name=MA0139.1_danRer11_dna-;Alias=CTCF;ID=MA0139.1-CTCF-1-danRer11_dna;pvalue=4.22e-06;qvalue= 0.14;sequence=TTTCCACTAGACGGCACAA;
danRer11_dna    fimo    nucleotide_motif    19573    19591    50.8    +    .    Name=MA0139.1_danRer11_dna+;Alias=CTCF;ID=MA0139.1-CTCF-2-danRer11_dna;pvalue=8.24e-06;qvalue= 0.14;sequence=agtccagaagagggcagat;

SEQUENCE INPUT:

>danRer11_dna range=chr9:56249111-56274247 
caggggtgcccaaactttttctcataaatggccaaaaaccaaacttgattgtgggctaaagttaaatataccagaccatattacctttagtttgccatac

So...how do I get the coordinates from my input (chromosome and base position) to appear in the output?
Then...how do I get the GFF output uploaded as a custom track onto the genome in UCSC?

[CharlesEGrant]

SEQUENCE INPUT:
>danRer11_dna range=chr9:56249111-56274247 
caggggtgcccaaactttttctcataaatggccaaaaaccaaacttgattgtgggctaaagttaaatataccagaccatattacctttagtttgccatac

If you are using the command line version of FIMO you can use the '--parse-genomic-coord' option. Unfortunately it looks like the description of this option was omitted from the current documentation. However it is described in the older version available here. Note that there is no accepted standard for encoding genomic coordinates in FASTA sequence headers. FIMO's '--parse-genomic-coord' is only able to extract coordinates from the USCC style headers. The sequence header for your sample would have to changes to 

>chr9:56249111-56274247

We'll try to have a patch fixing the current documentation shortly.