Custom features in IGV

Question

[jhanks1981]

I am making gtfs file with custom feature types, and I want to inspect them in IGV viewer. However, I cannot view the features if they have a non-standard attribute type (INTRON). However, if I replace the string "INTRON" with "exon", the file displays as expected. Is IGV viewer unable to display anything with a non-standard attribute name, of is there something I am missing? Here are the first two files that demonstrate the issue:

    $ head -5 22_exons.gtf
    chr22   my_calculation  exon    16062326        16062801        .       +       .       gene_id "ENSG00000233866.1"; transcript_id "ENST00000424770.1"; exon_number 1.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
    chr22   my_calculation  exon    16100658        16101266        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 3.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
    chr22   my_calculation  exon    16101484        16118782        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 2.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
    chr22   my_calculation  exon    16118923        16122710        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 1.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
    chr22   my_calculation  exon    16123778        16124732        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 1.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";

And this is the same file, except that "exon" is changed to "INTRON"

$ head -5 22_introns.gtf
chr22   my_calculation  INTRON  16062326        16062801        .       +       .       gene_id "ENSG00000233866.1"; transcript_id "ENST00000424770.1"; exon_number 1.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
chr22   my_calculation  INTRON  16100658        16101266        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 3.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
chr22   my_calculation  INTRON  16101484        16118782        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 2.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
chr22   my_calculation  INTRON  16118923        16122710        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 1.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";
chr22   my_calculation  INTRON  16123778        16124732        .       -       .       gene_id "ENSG00000223875.1"; transcript_id "ENST00000420638.1"; exon_number 1.5; exon_id "no_value"; ccdsid "no_value"; gene_name "no_value"; gene_status "no_value"; gene_type "no_value"; havana_gene "no_value"; havana_transcript "no_value"; level "no_value"; ont "no_value"; tag "no_value"; transcript_name "no_value"; transcript_status "no_value"; transcript_type "no_value";

and here is a picture of what I see in IGV with the two files:

https://ibb.co/bCnmP5

Additional details: I am just adding stuff because the forum apparently did not recognize my post as the English language. There is no more relevant information after this. I am just adding sentences because I read that it can help me get past the spam detector. Please forgive this irrelevant stuff. Wow, what an overly aggressive spam detector. My grammar isn't that terrible, is it?

[GenoMax]

You have not attached a link for the image in your post. Can you post a direct link in the body of the text?

Custo, features in IGV

Is that some random text in the title as well?

[jhanks1981]

Sorry, the link was in the raw post, but for some reason it didn't show. The "Custo," was supposed to be "Custom".

[Istvan Albert]

It is not the grammar just too many other types of words in the data I would guess.

[Istvan Albert]

I tried your example since I found that surprising and I often have non-standard feature names. For me it does draw lines even when I have the word INTRON as a type.

But IGV can be finicky and there are many undocumented features across versions.

[jhanks1981]

Interesting. I am on version 2.3.97 for Linux. Is there a better/more consistent viewer than IGV?

[Istvan Albert]

I would double check that file - it really does not make sense for IGV to now show a feature based on type alone.

I can reason for drawing some differently (CDS are thicker etc) but I doubt that it would be desirable to hide a valid interval from sight. I suspect file format error. Or host your file somewhere and we can check that ourselves.

Answer 1

[Dan D]

The IGV documentation references the official spec for GTF2. That spec only allows for the following values:

• CDS
• start_codon
• stop_codon
• exon

Granted, the GTF format is a special-case derivation of GFF so that's why it doesn't allow introns as a feature type. I'd think you would be fine if you made the appropriate adjustments and loaded the file as a GFF. Have you tried that?

[jhanks1981]

I will try that. Strange though, I've used GTF files with other feature types ("gene", "transcript") without issue. Also, I was reading about the GTF format from here: http://www.ensembl.org/info/website/upload/gff.html

Edit: Yes, that did the trick. Thanks!