15.15 : Genome Annotation and Assembly

The entire genome of an organism cannot be sequenced in continuous sequences - even the newest generation of sequencing technologies produce fragmented data from thousands of short DNA fragments ranging from 50-1000bp in length.

These short DNA sequences - called reads - need to be assembled to reconstruct the complete sequence of a genome in a process called genome assembly.

There are four main steps in any next-generation genome assembly - raw data analysis, contig assembly, scaffolding, and finally, gap closing.

The first step is to analyze the raw data acquired for quality - and then eliminate any contamination, biased data, or poor quality reads with a large number of unknown nucleotides.

Next, the clean reads are trimmed to remove the adapter sequences from their ends. Any bases at the fragment ends that do not pass the quality threshold are also trimmed.

Then, a well-suited assembly tool is used to assemble the reads into contiguous sequences - called contigs - based on the overlapping DNA segments.

Comparative genome assembly can be used when a reference genome of a closely related organism is available to direct the reconstruction of the new genome. Here, the reads are aligned to the reference genome, and this provides a layout for further steps in the genome assembly.

Alternatively, de novo genome assembly needs to be performed in the absence of a reference genome. Here, the overlapping reads are used to orient the sequences into longer contigs.

In the next step, the paired short reads - which are overhanging reads at the end of the contigs -  are used for scaffolding the genome.

The gaps between the adjacent contigs are filled with Ns incase of unknown sequences. However, if long reads that are more than 1kb in length are used to stitch contigs together, the gaps can be filled with actual sequences.

The result is an assembled genome that then needs to be annotated with the help of automated tools - a process called genome annotation.

The two main aims of genome annotation are gene structure and gene function prediction, commonly known as structural annotation and functional annotation, respectively.

While the structural annotation leads to the identification of the genomic elements such as coding regions, regulatory motifs,etc.; the functional annotation helps to correctly identify the biological function of these structural elements especially, protein-coding genes.

Genome annotation tools use available data, including known transcripts, protein or signal sequences, predicted genes from other sequenced genomes, or signatures of conserved domains, as the references for any new annotation.

Once the software aligns this available data to the draft genome, it needs to be filtered and polished either manually or using annotation tools to obtain a final set of gene annotations.