Element Biosciences

Company

The company was founded in 2017 [1] and has raised >95MUSD, from Fidelity, Venrock, JSCapital, and foresight capital (among others?).

The founding team is all ex-Illumina. The CEO (Molly He) was Senior Director of Protein Engineering and Enzymology at Illumina. The co-founding CTO (Michael Previte) was Associate Principal Scientist at Illumina. And co-founder Matt Kellinger was Staff Scientist at Illumina. There are currently 69 employees listed on LinkedIn.

Technology

There’s not much publicly disclosed on Element’s technological approach. However we can get some idea from patents and job adverts. Job adverts show positions in “Optical and Systems Engineering” and “Image and System Processing”.

They also have a couple of patents published, one refers to “Low non-specific binding supports and formulations for performing solid-phase nucleic acid hybridization and amplification” [3].

Essentially this patent refers to a process for creating a surface which oligos can be attached to. This serves a similar function to the flowcell surface in Illumina sequencing. The surface described appears to be compatible with clonal amplification.

So essentially this looks very similar to the Illumina/Solexa approach. A surface, to which oligos are attached, which undergo cluster growth (clonal amplification) [5]. Sequencing could then occur using sequencing-by-synthesis.

The patent claims better CNR (contrast-to-noise ratio), than other surfaces. And shows a figure comparing various surfaces:

These plots look similar to the pairwise plots you see when comparing the overlapping emissions in Illumina sequencing [4], in particular these plots look similar to the AC pairwise intensity plots. The Reference surface looks pretty bad (which appears to be a commercially available surface), the high CNR surface looks broadly similar to what I’ve seen in the past (Genome Analyzer 2 data [4]).

Improved CNR is clearly of benefit. However in terms of data quality, other factors likely limit accuracy and read length. If improved CNR allows for shorter imaging times/less illumination this could help with photo-bleaching issues. However it seems likely that phasing would still be a issue and limit read lengths.

Overall, going by the patent and job adverts, my best guess would be that Element are building an optical (likely not single molecule) sequencing-by-synthesis platform. As such, projected accuracy and read length would be in the same ballpark as Illumina. Illumina’s margins are quite high and there’s certainly room to compete on pricing. Many of the original Solexa patents are expiring, so they probably have freedom to use the same basic clonal amplification/sequencing-by-synthesis approach. I would guess that there are also reversible terminators that are available to them (or they could just do without).

Notes

[1] According to Crunchbase. https://www.crunchbase.com/organization/element-bioscience#section-overview There website lists a 80.3MUSD round on Jan 9th 2020. And a total of “more than $100 million”. They also list a series A of 15MUSD on July 19th 2019. This would suggest that sometime prior to the series A they raised ~5MUSD. Which would make sense as a large seed round.

[2] https://jobs.lever.co/elembio

[3] http://www.freepatentsonline.com/y2020/0149095.html

[4] See https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2734321/ figure 2.

[5] There are a number of “spotty” images in the patent. These look broadly similar to the random cluster images you’d see from Genome Analyzer 2 era Illumina instruments.