Back

A benchmark of Hi-C scaffolders using reference genomes and de novo assemblies

Sur, A.; Noble, W. S.; Myler, P. J.

2022-04-20 genomics
10.1101/2022.04.20.488415 bioRxiv
Show abstract

BackgroundStudying a new species using high-throughput sequencing requires a high-quality reference genome. However, assembling chromosome length sequences remains challenging. Recent advances in chromatin conformation capture (Hi-C) have provided a new approach to scaffolding genome assemblies, and the last ten years have seen a proliferation of such methods. However, to our knowledge no comprehensive benchmarking of Hi-C scaffolders has been conducted to date. ResultsThrough a literature review we identified the most popular Hi-C scaffolders - Lachesis, HiRise, 3d-dna, SALSA, and AllHiC. We tested their ability to scaffold four well studied genomes - S. cerevisiae, L. tarentolae, A. thaliana, and H. sapiens. Scaffolders were tasked with both scaffolding fragmented versions of the reference genome as well as de novo assemblies derived from long read datasets. We found that all scaffolders can exceed 80% accuracy under ideal circumstances but that their performance quickly deteriorates under more challenging conditions. Surprisingly, many scaffolders also showed poor performance on the best assemblies, where contigs are near chromosome length. Overall, we found that HiRise and Lachesis offer the best performance on average across all conditions. ConclusionsWe compare the performance of five Hi-C scaffolders using multiple reference species under both ideal and real-life conditions, thereby illuminating their strengths and weaknesses.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.