2.5 GTF Format

GTF (Gene Transfer Format) is a widely-used file format for gene annotations. It’s a refinement of GFF version 2 and is particularly popular for RNA-seq analysis.

Overview

GTF files describe:

Gene structures
Transcript isoforms
Exon positions
Coding sequences (CDS)
Start and stop codons

GTF Format Structure

GTF files are tab-delimited text files with 9 columns, similar to GFF but with specific attribute requirements.

Basic Structure

chr1    HAVANA  gene    11869   14409   .   +   .   gene_id "ENSG00000223972"; gene_name "DDX11L1";
chr1    HAVANA  transcript  11869   14409   .   +   .   gene_id "ENSG00000223972"; transcript_id "ENST00000456328"; gene_name "DDX11L1";
chr1    HAVANA  exon    11869   12227   .   +   .   gene_id "ENSG00000223972"; transcript_id "ENST00000456328"; exon_number "1";
chr1    HAVANA  exon    12613   12721   .   +   .   gene_id "ENSG00000223972"; transcript_id "ENST00000456328"; exon_number "2";
chr1    HAVANA  exon    13221   14409   .   +   .   gene_id "ENSG00000223972"; transcript_id "ENST00000456328"; exon_number "3";

GTF Columns

Column	Name	Description
1	seqname	Chromosome or scaffold name
2	source	Source of annotation (Ensembl, HAVANA, etc.)
3	feature	Feature type (gene, transcript, exon, CDS, etc.)
4	start	Start position (1-based, inclusive)
5	end	End position (1-based, inclusive)
6	score	Confidence score (usually .)
7	strand	+ (forward), - (reverse)
8	frame	For CDS: 0, 1, or 2 (reading frame)
9	attribute	Semicolon-separated key “value” pairs

Standard feature types:

gene - Gene
transcript - Transcript (mRNA, lncRNA, etc.)
exon - Exon
CDS - Coding sequence
start_codon - Translation start
stop_codon - Translation stop
UTR - Untranslated region
5UTR - 5’ UTR
3UTR - 3’ UTR

Reading frame

For CDS features - reading frame:

0 - First nucleotide is first position of codon
1 - Second nucleotide is first position of codon
2 - Third nucleotide is first position of codon
. - Not applicable (for non-CDS features)

Attributes

Required attributes:

gene_id - Unique gene identifier (REQUIRED for all features)
transcript_id - Unique transcript identifier (REQUIRED for transcript-level features)

Attribute format:

key "value"; key2 "value2";

Note: Values are quoted, statements end with semicolon

Example: Complete Gene Structure


# Transcript level
chr11   refGene transcript      32233672        32234486        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218";  gene_name "Hba-a1";

# Exons
chr11   refGene exon    32233672        32233801        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "1"; exon_id "NM_008218.1"; gene_name "Hba-a1";
chr11   refGene exon    32233924        32234128        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "2"; exon_id "NM_008218.2"; gene_name "Hba-a1";
chr11   refGene exon    32234263        32234486        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "3"; exon_id "NM_008218.3"; gene_name "Hba-a1";

Example: Protein-Coding Gene (Hba-a1)

chr11   refGene transcript      32233672        32234486        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218";  gene_name "Hba-a1";
chr11   refGene exon    32233672        32233801        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "1"; exon_id "NM_008218.1"; gene_name "Hba-a1";
chr11   refGene 5UTR    32233672        32233706        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "1"; exon_id "NM_008218.1"; gene_name "Hba-a1";
chr11   refGene CDS     32233707        32233801        .       +       0       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "1"; exon_id "NM_008218.1"; gene_name "Hba-a1";
chr11   refGene exon    32233924        32234128        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "2"; exon_id "NM_008218.2"; gene_name "Hba-a1";
chr11   refGene CDS     32233924        32234128        .       +       1       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "2"; exon_id "NM_008218.2"; gene_name "Hba-a1";
chr11   refGene exon    32234263        32234486        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "3"; exon_id "NM_008218.3"; gene_name "Hba-a1";
chr11   refGene CDS     32234263        32234388        .       +       0       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "3"; exon_id "NM_008218.3"; gene_name "Hba-a1";
chr11   refGene 3UTR    32234392        32234486        .       +       .       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "3"; exon_id "NM_008218.3"; gene_name "Hba-a1";
chr11   refGene start_codon     32233707        32233709        .       +       0       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "1"; exon_id "NM_008218.1"; gene_name "Hba-a1";
chr11   refGene stop_codon      32234389        32234391        .       +       0       gene_id "Hba-a1"; transcript_id "NM_008218"; exon_number "3"; exon_id "NM_008218.3"; gene_name "Hba-a1";

Working with GTF Files

View GTF file

less annotation.gtf
head -n 20 annotation.gtf

Count features by type

cut -f3 annotation.gtf | sort | uniq -c

Extract all 5’UTRs

awk '$3 == "5UTR"' annotation.gtf

Extract specific gene

grep 'gene_name "ACTB"' annotation.gtf

Get unique gene list

grep -P '\tgene\t' annotation.gtf | \
    grep -oP 'gene_name "\K[^"]+' | \
    sort -u

Extract transcripts on specific chromosome

awk '$1 == "chr1" && $3 == "transcript"' annotation.gtf

Convert GTF to BED

# Extract exons as BED
awk '$3 == "exon" {print $1"\t"$4-1"\t"$5"\t"$10"\t"$6"\t"$7}' annotation.gtf | \
    tr -d '";'

Gene GTF Sources

Internal to the CCB:

/databank/igenomes/Mus_musculus/UCSC/mm39/Annotation/Genes/refGene.gtf

/databank/igenomes/Homo_sapiens/UCSC/hg38/Annotation/Genes/genes.gtf