Back

Whole genome sequencing with AVITI and NovaSeq X Plus reveals comparable performance with contextual biases

Höjer, P.; Alneberg, J.; Lundin, P.; Martin, T.; Hauenstein, J.; Fällmar, H.; Lindell, M.; Natanaelsson, C.; Häggqvist, S.; Ameur, A.; Nordlund, J.; Mansson Welinder, R.

2025-10-10 genomics
10.1101/2025.10.10.681584 bioRxiv
Show abstract

Element Biosciences avidity sequencing has emerged as a competing technology to Illuminas short read sequencing platform. Prior benchmarks of avidity sequencing have not included the latest Illumina NovaSeq X/X Plus instruments with XLEAP chemistry. Here, we have run PCR-free whole genome sequencing on four human tumor cell lines using both Illumina NovaSeq X Plus and Element AVITI instruments. AVITI showed low duplication rates and reported higher base qualities, the latter contributed to improved mapping confidence and fewer spurious variant candidates. Both platforms were found to be highly comparable when benchmarking variant calling, with AVITI only providing a minor improvement on INDELs at lower coverages. Stratifying by genomic context revealed further differences, where AVITI genome coverage and variant calls were superior in high GC regions while being inferior in GC homopolymers. Error rate analysis highlighted further differences between the platforms, in particular AVITI in some instances displayed an increased error rate on read 2 related to short fragments. AVITI error rate was also found to be more stable downstream of repetitive regions, except for GC homopolymers. We further found that AVITI sequencing was sensitive to G-quadruplex motifs. Overall, despite these identified differences, both platforms performed highly comparable for variant analysis.

Published in NAR Genomics and Bioinformatics (predicted rank #3) · training set

Matching journals

The top 3 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.