Glossary - hillerlab/TOGA2 GitHub Wiki

Assembly gap

A genomic locus of unknown sequence in the assembly, typically represented by a stretch of ambiguous bases (N's).

Chain

A chain of co-linear local alignments consist of local sequence alignments that occur in the same order and strand in the reference and the query genome. Local alignments can be separated by insertions/deletions in one species, or regions with unaligning sequence.

Compensation

Referring to two or more frameshifts that together restore the reference reading frame. For example, a -1 bp deletions followed by a +1 insertion without leading to a stop codon in the sequence affected by the shifted frame. Another example are three -1 bp deletions that together result in a -3 bp (non-frameshifting) deletion.

Exon

Part of a gene that remains in the mature RNA transcript after splicing. Note that exons can consist of coding or untranslated regions, meaning some exons can be entirely part of the 5' or 3' untranslated region.

Gene

TOGA2 considers only coding genes. A DNA sequence that contains the information to produce a protein via translation.

Inactivating mutation

A mutation that disrupts the reading frame or splice site. Specifically, inactivating mutations considered by TOGA2 are frameshifting deletions and insertions, stop codon mutations, exon deletions, disruption of the splice site dinucleotides.

Indel

A joint term for insertions and deletions.

Isoform

Same as Transcript [2]; a distinct type of transcripts encoded by a given gene.

Isoform file

A two-column file containing gene-to-transcripts mapping. Each line corresponds to a single gene-transcript pair (the so-called ‘long data’ format). Generally, the term refers to gene-transcript mapping for the reference genome assembly, provided to TOGA2 with --isoform_file argument. A counterpart for the resulting query annotation is query_genes.tsv file.

Loss status

The loss status provides information about the intactness of the reading frame and the presence of exons. Simply put the loss status provides information about whether a transcript likely encodes a functional protein (intact reading frame), has inactivating mutations that likely prevent translation into a functional protein, or has missing sequences (often due to assembly gaps or assembly fragmentation) such that the transcript is not fully present in the genome assembly.

Masked mutation

A mutation that is recognized by TOGA2 but not considered to be gene-inactivating. Such mutations are not taken into account when inferring the loss status. Specifically, TOGA2 masks the following mutations: start/stop codon mutations, U12 or non-canonical U2 splice sites, compensated frameshifts (e.g. a -1 bp del followed by a +1 insertion without leading to a stop codon in the sequence affected by the shifted frame), precise intron deletions, and new query introns.

Ortholog

A homologous gene (having a shared ancestry) that originated by a speciation event.

Orthologous chain

A co-linear alignment chain that aligns a transcript in the reference to an orthologous locus in the query genome.

Orthologous projection

Refers to a transcript of a reference gene that is aligned to an orthologous query locus with an orthologous chain.

Paralog

A homologous gene (having a shared ancestry) that originated by a duplication event.

Paralogous chain

A co-linear alignment chain that aligns a transcript in the reference to a paralogous locus in the query genome.

Paralogous projection

Refers to a transcript of a reference gene that is aligned to a paralogous query locus with an paralogous chain.

Processed pseudogene

A reverse transcribed, spliced mRNA that was integrated back into the genome. Since processed pseudogene lack the promoter (because the promoter is not transcribed), they are often 'dead on arrival' in the new locus. However, processed pseudogenes can be transcribed using existing promoters in the locus, and they sometimes evolve into new genes.

Projection

A projection refers to a transcript of a reference gene that corresponds to a locus in the query genome. Projections are the core unit of TOGA2 annotations.

Retrogene candidate

A processed pseudogene that encodes an intact reading frame and thus has the potential capacity to encode a functional gene. TOGA2 annotates such retrogene candidates, as they are candidates for new genes in the query if they are transcribed.

Second-best projection

Refers to cases where several projections of the reference transcripts belonging to different genes align to the same query locus. The projections can be ranked by their orthology probability (determined in TOGA's machine learning classifier). The top projection is the one with the highest orthology probability; all other projections are referred to as second-best projections in this query locus. TOGA2 uses this information to discard weak (often wrong) orthologs at the query locus.

TOGA

A Tool to infer Orthologs from Genome Alignments. Depending on the context, the name may refer specifically to the original TOGA tool, but generally is used for both TOGA and TOGA2.

TOGA1

The original version of TOGA published in 2023 (DOI: 10.1126/science.abn31)

TOGA2

The direct successor of TOGA(1), hosted in this repository.

Transcript

1. (Generally) A single RNA molecule produced during the transcription process, or
2. (TOGA2 terminology) Same as transcript isoform: a distinct type of transcripts encoded by a given gene, or
3. (TOGA2 terminology) reference transcript isoform; query transcripts are usually referred to as projections