BlobToolKit in the real world 2: Horizontal gene transfer?

In November 2015, an American university published a paper reporting that they had sequenced the first ever complete genome of a tardigrade species known as Hypsibius dujardini.

Tardigrades are miniscule animals that are closely related to Nematodes (roundworms) and Arthropods (creatures with hard outer bodies and jointed limbs, including insects, spiders, crustaceans, etc). The paper claimed that as much as 17% of the tardigrade’s DNA was horizontally transferred from bacteria and other microorganisms (Horizontal Gene Transfer: When genes or genetic information is transferred “horizontally”, eg, from one organism to another, as opposed to “vertically” where the genetic information is transferred from a parent to an offspring via reproduction). This was an extraordinary claim as the previous highest estimate for horizontal gene transfer (HGT) in animals was only 2-3%.

When this study was published, two of the instructors on this Future Learn course (Mark Blaxter and Sujai Kumar) had been working on the same species in a Scottish university and we had not seen any evidence for this horizontally transferred DNA. We reanalysed their data which they kindly shared in the spirit of open science. We realised very quickly that the American group had not been able to check for contaminants in their tardigrade sample.

When we plotted their own data as BTK plots – we quickly saw that there were several contaminants present at different coverages, and they showed up as separate blobs.

We published our findings and the genomics research community agreed that our interpretation of the data was more accurate. BlobToolKit was able to detect the presence of contaminating bacteria and other organisms and confirm that tardigrades do not have extensive horizontal gene transfer from bacteria.

In this article, we want to show you what the original incorrect Hypsibius dujardini genome assembly looks like in BTK, and what the final reference genome for that species looks like.

BTK view of original incorrect Hypsibius dujardini genome assembly – LMYF01

BTK view of final reference Hypsibius dujardini genome assembly – MTYJ01

In the original incorrect LMYF01 genome assembly you can see a large central blob centred around a coverage of 10 and GC of 0.45 (or 45%). In addition there are several large light green (Proteobacteria) and dark green (Bacteroidetes) circles, as well as some orange and green and purple circles towards the lower right of the plot. All of these are obvious contaminants as they don’t share the same concentration as the main central blob and have consistently different taxonomic labels.

In the final MTYJ01 assembly, there is only one main blob. There is a lower light-blue blob with many small circles representing a small proportion of the genome assembly at lower coverage. These circles are likely to be small-scale contaminants, but nowhere near the scale of the large bacterial contigs in the previous plot.

We will discuss these issues in more detail later on in the course, to understand how the taxonomic assignments are done in BTK. For now we just wanted you to see what a heavily contaminated genome assembly looks like in the BTK viewer.