Genome annotation

Protein-coding gene reconstruction focuses on identifying and annotating the locations of protein-coding genes within a DNA sequence. In the vast landscape of genomic information, the accurate prediction of structural genes is crucial for unraveling the functional elements that govern an organism's biology.

Structural gene prediction takes into account various features, such as open reading frames (ORFs), splice sites, start and stop codons, and consensus sequences, to distinguish coding regions from non-coding ones.

We use a newly developed, proprietary algorithm for homology-based gene reconstruction. To adapt the homology model, we created thousands of genome annotations internally and improved them iteratively.

Functional gene annotation involves assigning biological functions to the discovered genes within a genome. We assign among others

domain architecture
Domain profiles from Pfam, CDD, ProSitePattern and others.
protein/gene names
Protein naming strongly depends on the reference database used and the quality and completeness of the query sequence. In addition, naming subfamily members within protein families is error prone. In addition to potentially misleading subfamily designations we apply protein family names.
GO terms
GO-terms are derived and combined from gene names and domain profiles.
EC numbers
EC numbers (Enzyme Commission numbers) are a numerical classification scheme for enzymes, based on the chemical reactions they catalyze

tRNA annotation types include

cognate
anticodon matches the tRNA isotype; no pseudo genes
non-cognate isotype
anticodon does not match the tRNA isotype
pseudo
score of the tRNA-sequence match too low
undetermined
anticodon contains ‘N’
truncated
tRNA not complete, sequence at either or both ends missing
suppressor
anticodon matches one of the stop codons

Predicted RNA gene families includ ribosomal RNA, spliceosomal RNA, RNase P, RNase MRP, telomerase and snoRNA U3 genes

The exon regions upstream (5') and downstream (3') of the coding exon regions are called untranslated regions (UTRs), 5'UTRs (also known as leader sequences) and 3'UTRs (also known as trailer sequences) respectively. UTRs can be interrupted by introns or alternatively spliced like all other exon regions.

UTRs are determined by mapping RNASeq data or by prediction. Predictions are made from profiles trained with UTRs based on RNASeq data or from profiles of closely related species.

Coverage by RNASeq data can be defined in two ways: i) Horizontal coverage, i.e. the support of the gene model from the 5’ to the 3' end. In this case, the total length of the exons is defined as 100% and the coverage is the percentage of mRNA covered by the RNASeq data. ii) Vertical coverage, i.e. the support of each nucleotide by RNASeq reads.

Here we use the vertical coverage to indicate whether a feature is supported by experimental data. The value indicates whether there is strong support (high values) or low support (low values). Note: The values must not be used for differential expression analysis!

Due to their similarity to genes and gene elements, retrotransposons are incorrectly predicted to a considerable extent by common gene prediction programmes. Depending on the genome, 5 to 10 % of the genome annotations in public databases consist of transposons and/or contain parts of transposons in their gene structure.

We use a novel approach to unambiguously identify transposons and transposon-fragments. By annotating all types of genome features, protein-coding, non-coding and RNA-genes, pseudogenes and transposons, we ensure that one feature is not mispredicted as another.

You will select 5 genes for final annotation, which we will analyse manually. This includes the confirmation of exons by RNA-Seq data or comparative genomics and the annotation of the most important alternative splice variants.

The full functional annotation as well as RNA-Seq evidence, if available, are presented as separate sections in the annotation report.

The data analysis contains additional, biology-based analyses to the report including many interactive charts and plots. Biology-based analyses include, for example, the mapping tRNA genes onto the genetic code, a detailed splice-pattern analysis, a detailed functional annotation analysis and many more.

assembly evaluation
N50 plot, A50 plot, GC content per contig plot
genome annotation metrics
gene stats, exon/intron lengths statistics
distribution of genes on strands
genome covered by genes/features
codon usage frequency
exon phase distribution
annotation quality indicators
intron type distribution (3n/3n+1/3n+2)
analysis of introns with respect to in-frame stop codons 
splice site pattern distribution

only in combination with UTR package
distribution of genes with UTRs
genes covered by RNA-Seq