BlobToolKit - Interactive Quality Assessment of Genome Assemblies.
Challis R, Richards E, Rajan J, Cochrane G, Blaxter M. · 2020
Read BlobToolKit - Interactive Quality Assessment of Genome Assemblies. by Challis R, Richards E, Rajan J, Cochrane G, Blaxter M. (2020) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc7144090.txt.
Introduction
Challis and co-authors present BlobToolKit, software for spotting contaminant DNA in genome assemblies, in G3: Genes, Genomes, Genetics. Sequencers cannot tell which species a DNA fragment came from, so stray DNA from handling or co-extraction can end up inside a published genome. Such mixtures mislead later studies.
The toolkit processes assembly and read files into interactive browser plots that separate target DNA from foreign sequences. It can filter assemblies during production and audit public ones. The team runs it automatically on eukaryotic genomes in the public archives and posts the results in an open viewer, with quality views being linked from the European Nucleotide Archive record pages.
Catalog record
| CREATOR | Challis R, Richards E, Rajan J, Cochrane G, Blaxter M. |
|---|---|
| YEAR | 2020 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
BlobToolKit – Interactive Quality Assessment of Genome Assemblies By Richard Challis, Edward Richards, Jeena Rajan, Guy Cochrane, Mark Blaxter Published in: G3: Genes|Genomes|Genetics DOI: 10.1534/g3.119.400908 ABSTRACT Abstract Reconstruction of target genomes from sequence data produced by instruments that are agnostic as to the species-of-origin may be confounded by contaminant DNA. Whether introduced during sample processing or through co-extraction alongside the target DNA, if insufficient care is taken during the assembly process, the final assembled genome may be a mixture of data from several species. Such assemblies can confound sequence-based biological inference and, when deposited in public databases, may be included in downstream analyses by users unaware of underlying problems. We present BlobToolKit, a software suite to aid researchers in identifying and isolating non-target data in draft and publicly available genome assemblies. BlobToolKit can be used to process assembly, read and analysis files for fully reproducible interactive exploration in the browser-based Viewer. BlobToolKit can be used during assembly to filter non-target DNA, helping researchers produce assemblies with high biological credibility. We have been running an automated BlobToolKit pipeline on eukaryotic assemblies publicly available in the International Nucleotide Sequence Data Collaboration and are making the results available through a public instance of the Viewer at https://blobtoolkit.genomehubs.org/view. We aim to complete analysis of all publicly available genomes and then maintain currency with the flow of new genomes. We have worked to embed these views into the presentation of genome assemblies at the European Nucleotide Archive, providing an indication of assembly quality alongside the public record with links out to allow full exploration in the Viewer. Keywords: Bioinformatics, visualisation web-tool, genome assembly, quality control Genome sequences are part of the basic data economy of modern bioscience. Using assembled genomes, it is possible to identify loci underpinning key traits of interest, discover the regulatory logic of gene expression, investigate disease processes, and explore the evolutionary histories of genes and species. These research programs rely implicitly on the correctness of the genome sequences. Errors in genome sequences risk distracting or even derailing their effective use. Assembly of true genome sequences from reads shorter than the length of a replicon remains a difficult task (Ekblom and Wolf 2014). This task is made more complex when isolation of the original samples or the processing of DNA to generate the raw sequence data cannot avoid contamination of the target genome with DNA from non-target sources (Salzberg et al. 2005; Salter et al. 2014). Sequencing instruments are agnostic as to species-of-origin of the fragments they are tasked with processing, and thus a contaminated sample will result in a contaminated raw dataset. If insufficient care is taken during the assembly process, this can mean that the final assembled genome is a mixture of data from several species, and cannot be used as a good representation of the target species (Merchant et al. 2014). Downstream, this can result in erroneous attribution of biochemical or genetic properties to the target species that are actually derived from the contaminants’ genomes (Artamonova et al. 2015; Arakawa 2016). However, not all “contaminants” are uninteresting. Many eukaryotic species live in close biological association with symbionts, and many bacteria exist in, and can only be grown as, consortia of interacting species (López-García et al. 2017). In these systems genome sequencing aims to reconstruct the genomes of all the independent species and strains involved (Kumar and Blaxter 2011). We are developing BlobToolKit, a software suite that will aid researchers in identifying contamination before it is erroneously blessed as being part of a target genome and to separate sequences that belong to different members of biological consortia. BlobToolKit is based on BlobTools written by Dominik Laetsch (Laetsch and Blaxter 2017) which was in turn based on the original Blobology pipeline from Sujai Kumar (Kumar et al. 2013). We present a toolkit that has been rewritten in its entirety to make use of advanced web frameworks and visualization. Like its progenitors, BlobToolKit uses GC proportion and coverage as two major axes on which contigs or scaffolds from an assembly can be displayed and overlays a taxonomic assignment to help separate the signal from noise when using these measures to identify sequence from different sources. Genomes from different taxa frequently have different mean GC proportion, with values ranging from 34–50% among monocot species (Šmarda et al. 2014), 44–58% among mammals (Romiguier et al. 2010) and 16–75% among bacteria (Lightfield et al. 2011). Intra-genomic GC proportion typically shows a unimodal distribution around a mean value, particularly in bacteria (Bohlin et al. 2010), however, regions of differing composition may yield bi- (or multi-) modal distributions (Salinas et al. 1988). These intragenomic differences result from biological processes, such as biased gene conversion (Galtier et al. 2001), and may be as large or larger than differences between species (e.g., 35–60% in the human genome (Lander et al. 2001)). Primary separation of sequences from different sources may be possible on this axis, provided the organisms have distinct mean values with relatively little deviation. The read coverage of each contig or scaffold in an assembly is an estimate of the relative stoichiometry of the replicon from which it derives. All the contigs or scaffolds from one species should have similar coverage, however, as with GC content, there are a number of confounding factors to consider. Organelles will usually have high coverage relative to the nuclear genome (Bakker et al. 2016) and collapsed segmental duplications in the assembly will also have elevated coverage (Bailey et al. 2002). Sex chromosomes are expected to have 50% coverage in the heterogametic sex (Tomaszkiewicz et al. 2017) and uncollapsed haploid segments of autosomes will also have 50% coverage. Contaminant or cobiont genomes are likely to have different, internally consistent stoichiometry, in which case they may be distinguished on this axis. To provide initial identification of sets of contigs or scaffolds from distinct taxa, BlobToolKit also decorates each scaffold or contig with a taxonomic attribution. This is assessed using the sequence similarity search tools BLAST (Altschul 1997) and Diamond (Buchfink et al. 2015) to perform local alignments against public sequence databases to identify the most closely matching sequences. Two tools are used as BLAST is most efficient at identifying closely related sequences in nucleotide databases, while Diamond is more efficient when searching against protein databases (Buchfink et al. 2015). This taxonomic attribution is tentative due to the presence of mis-annotated records in the public databases. In conjunction with GC proportion and coverage measures this serves to highlight clusters (or blobs) of contigs that share distinct properties and coherent taxonomic source. This richly marked-up annotation of the assembly makes it possible to assess whether it derives from single or multiple source organisms. The BlobToolKit data can be used to separate contigs and scaffolds (and the reads that generated them) into separate bins for subsequent reanalyses. BlobToolKit can be used as part of the process of genome assembly, playing a role both in separating raw input data for assembly of distinct components and in quality assurance of the final product. For genome assemblies released publicly, BlobToolKit can be used to provide quality assurance and to identify issues that should be taken into consideration in downstream reuse of the data. Here we present the latest version of