GT-CSA(1)

💖 GT-CSA 💖 — the GenomeTools manual, but make it make sense

✨ NAME ✨

gt-csa — she takes spliced alignments out of a GFF3 file and turns them into consensus spliced alignments. That's it. That's the whole job. iconic behaviour, honestly 🦋

---

🌸 SYNOPSIS 🌸

gt csa [option ...] [GFF3_file]

---

🦋 DESCRIPTION 🦋

okay so bestie, here's the whole vibe: you hand her a GFF3 file, she looks at all your spliced alignments, and she clusters the ones that overlap enough to be considered the same thing — then she gives you back the consensus version. The details of "enough" are literally the first option down there. 👇

💅 OPTIONS 💅

• -join-length [value] — sets the join length used for the spliced alignment clustering. (default: 300) — this is the "how close do two alignments gotta be to get smooshed together" number. 💞
• -v [yes|no] — be verbose. (default: no) — say yes and she'll actually tell you what's going on. 🗣️
• -o [filename] — redirect output to the specified file. (default: undefined) — no file, no redirection, it just comes straight at you. 📄
• -gzip [yes|no] — write a gzip compressed output file. (default: no) 🎀
• -bzip2 [yes|no] — write a bzip2 compressed output file. (default: no) 🎀
• -force [yes|no] — force writing to the output file. (default: no) — she will NOT clobber stuff unless you say so, which is honestly so considerate of her. 🫶
• -help — display help and exit. 🆘
• -version — display version information and exit. 🏷️

---

🔥 EXAMPLE 🔥

Let's assume we have a GFF3 file called csa_example_spliced_alignments.gff3 and it contains these four overlapping spliced alignments — and bestie, they're represented as genes with exons as children, keep that in mind, it matters later:

##gff-version 3
##sequence-region   seq 1 290
seq .       gene    1       209     .       +       .       ID=gene1
seq .       exon    1       90      .       +       .       Parent=gene1
seq .       exon    110     190     .       +       .       Parent=gene1
seq .       exon    201     209     .       +       .       Parent=gene1
###
seq .       gene    1       290     .       +       .       ID=gene2
seq .       exon    1       90      .       +       .       Parent=gene2
seq .       exon    101     190     .       +       .       Parent=gene2
seq .       exon    201     290     .       +       .       Parent=gene2
###
seq .       gene    10      290     .       +       .       ID=gene3
seq .       exon    10      90      .       +       .       Parent=gene3
seq .       exon    110     190     .       +       .       Parent=gene3
seq .       exon    201     290     .       +       .       Parent=gene3
###
seq .       gene    181     290     .       +       .       ID=gene4
seq .       exon    181     190     .       +       .       Parent=gene4
seq .       exon    201     290     .       +       .       Parent=gene4
###

To compute the consensus spliced alignments we call:

$ gt csa csa_example_spliced_alignments.gff3

Which returns:

##gff-version 3
##sequence-region   seq 1 290
seq gt csa  gene    1       290     .       +       .       ID=gene1
seq gt csa  mRNA    1       290     .       +       .       ID=mRNA1;Parent=gene1
seq gt csa  exon    1       90      .       +       .       Parent=mRNA1
seq gt csa  exon    110     190     .       +       .       Parent=mRNA1
seq gt csa  exon    201     290     .       +       .       Parent=mRNA1
seq gt csa  mRNA    1       290     .       +       .       ID=mRNA2;Parent=gene1
seq gt csa  exon    1       90      .       +       .       Parent=mRNA2
seq gt csa  exon    101     190     .       +       .       Parent=mRNA2
seq gt csa  exon    201     290     .       +       .       Parent=mRNA2
###

As one can see, they've been combined into a consensus spliced alignment — and notice how she's now representing it as genes with mRNAs as children, which in turn have exons as children ✨ — with two alternative splice forms.

• The first and the third spliced alignment have been combined into the first alternative splice form (mRNA1)
• The second and the fourth spliced alignment into the second alternative splice form (mRNA2)

As one can see, the second exon from the first alternative splice form is shorter than the corresponding exon from the second alternative splice form. 👀 the way that detail is just sitting there being meaningful...

---

🙈 REPORTING BUGS 🙈

Report bugs to https://github.com/genometools/genometools/issues.

fr fr, if something's broken, tell the maintainers, not me — I'm just the hype woman. 💜

---

okay so that's the whole page, bestie — go cluster some alignments, you're gonna slay 🫶 This was the GenomeTools manual page for GT-CSA(1), from GenomeTools 1.6.6, dated 11/09/2025. periodt. 💅✨