Same transcript coordinates in gtf file, different transcript ID

Question

I have a gtf file from Ensembl, and I noticed that several "transcript" annotations have the exact same coordinates. See for instance the third and fourth transcripts ("Y74C9A.2b.1" and "Y74C9A.2b.4") for this gene:

$ grep "WBGene00022276" genes.gtf | awk '$3 == "transcript" {print}'
I   ensembl transcript  10413   16842   .   +   .   gene_biotype "protein_coding"; gene_id "WBGene00022276"; gene_name "nlp-40"; gene_source "ensembl"; gene_version "1"; p_id "P24752"; transcript_biotype "protein_coding"; transcript_id "Y74C9A.2a.2"; transcript_name "Y74C9A.2a.2"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS7921";
I   ensembl transcript  11495   16793   .   +   .   gene_biotype "protein_coding"; gene_id "WBGene00022276"; gene_name "nlp-40"; gene_source "ensembl"; gene_version "1"; p_id "P22572"; transcript_biotype "protein_coding"; transcript_id "Y74C9A.2b.2"; transcript_name "Y74C9A.2b.2"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS9220";
I   ensembl transcript  11499   16842   .   +   .   gene_biotype "protein_coding"; gene_id "WBGene00022276"; gene_name "nlp-40"; gene_source "ensembl"; gene_version "1"; p_id "P22572"; transcript_biotype "protein_coding"; transcript_id "Y74C9A.2b.1"; transcript_name "Y74C9A.2b.1"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS9215";
I   ensembl transcript  11499   16842   .   +   .   gene_biotype "protein_coding"; gene_id "WBGene00022276"; gene_name "nlp-40"; gene_source "ensembl"; gene_version "1"; p_id "P22572"; transcript_biotype "protein_coding"; transcript_id "Y74C9A.2b.4"; transcript_name "Y74C9A.2b.4"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS9215";
I   ensembl transcript  11505   16842   .   +   .   gene_biotype "protein_coding"; gene_id "WBGene00022276"; gene_name "nlp-40"; gene_source "ensembl"; gene_version "1"; p_id "P24752"; transcript_biotype "protein_coding"; transcript_id "Y74C9A.2a.3"; transcript_name "Y74C9A.2a.3"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS16820";
I   ensembl transcript  11618   16842   .   +   .   gene_biotype "protein_coding"; gene_id "WBGene00022276"; gene_name "nlp-40"; gene_source "ensembl"; gene_version "1"; p_id "P24752"; transcript_biotype "protein_coding"; transcript_id "Y74C9A.2a.1"; transcript_name "Y74C9A.2a.1"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS43845";
I   ensembl transcript  11623   16842   .   +   .   gene_biotype "protein_coding"; gene_id "WBGene00022276"; gene_name "nlp-40"; gene_source "ensembl"; gene_version "1"; p_id "P22572"; transcript_biotype "protein_coding"; transcript_id "Y74C9A.2b.3"; transcript_name "Y74C9A.2b.3"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS49800";

What does it mean? Do these two "transcripts" actually correspond to an exon, belonging to two different isoforms?

Answer 1

transcript objects cover the co-ordinates from the start of the first exon to the end of the last exon of a transcript (i.e. an isoform). If two different isoforms share the same first and last exons, but have a different set of internal exons, then their transcript entries will be the same, but the set of exon entires associated with each transcript will be different.

For example, consider the two transcripts below:

                      1         2
             12345678901234567890123 
transcript_1 |>>>|----|>>|----|>>>>|

transcript_2 |>>|--|>>|------|>>>>>|

These two transcripts would give the following GTF entires

chr1   ensembl   transcript    1   23   .   +   .   gene_id "gene_1"; transcript_id "transcript_1";
chr1   ensembl   exon    1   5   .   +   .   gene_id "gene_1"; transcript_id "transcript_1";
chr1   ensembl   exon    10   13   .   +   .   gene_id "gene_1"; gene_id "gene_1"; transcript_id "transcript_1";   
chr1   ensembl   exon    18   23   .   +   .   gene_id "gene_1"; transcript_id "transcript_1";   
chr1   ensembl   transcript    1   23   .   +   .   gene_id "gene_1"; transcript_id "transcript_2";
chr1   ensembl   exon    1   4   .   +   .   gene_id "gene_1"; transcript_id "transcript_2";
chr1   ensembl   exon    7   10   .   +   .   gene_id "gene_1"; transcript_id "transcript_2";   
chr1   ensembl   exon    17   23   .   +   .   gene_id "gene_1"; transcript_id "transcript_2";   

Answer 2

As Ian explained, these are different transcripts which happen to have the same start and end positions. You have no information on their exonic structure in that file. However, if you look them up at EnsEMBL, you will see:

Transcript Y74C9A.2b.1 :

1   Y74C9A.2b.1.e1  11,499  11,561  -   -   63
    Intron 1-2      11,562  11,617          56
2   Y74C9A.2b.1.e2  11,618  11,689  -   -   72
    Intron 2-3      11,690  14,950          3,261
3   Y74C9A.2b.1.e3  14,951  15,160  -   1   210
    Intron 3-4      15,161  16,472          1,312
4   Y74C9A.2a.1.e3  16,473  16,842  1   -   370

This has 4 exons, two of which are non-coding (UTR). The second starts at

Transcript Y74C9A.2b.4

1   Y74C9A.2b.4.e1  11,499  11,557  -   -   59
    Intron 1-2      11,558  11,617          60
2   Y74C9A.2b.1.e2  11,618  11,689  -   -   72
    Intron 2-3      11,690  14,950          3,261
3   Y74C9A.2b.1.e3  14,951  15,160  -   1   210
    Intron 3-4      15,161  16,472          1,312
4   Y74C9A.2a.1.e3  16,473  16,842  1   -   370

Granted, the two transcripts are very simlar, they both have 4 exons, the first 2 of which are non coding (UTR). However, the 1st exon of transcript Y74C9A.2b.1 is 63 nt long and ends at 11,561 while the 1st exon of transcript Y74C9A.2b.4 is 59 nts and ends at 11,557.

So there you go: two (slightly) different transcripts.