Scale Biosciences uses a combinatorial indexing, split-pool technique they've named Quantum Barcoding to enable single cell RNA-Seq at scales previously unavailable without microfluidic controllers. VizGen has one of the leading spatial RNA platforms with MERFISH.
The theme of the collaboration, as the CEOs put it, is "let's create some data". Scale and VizGen will be coordinating efforts to sponsor academic collaborations that demonstrate the value of each platform separately and which work together. The two companies will also co-market each others products; in effect each company has significantly expanded their sales team.
In a recent preprint, Austin Hartman and Rahul Satija outline one way for single cell and spatial to be used in a complementary fashion. In this case, they used single cell data as a check on spatial data and particularly around cell segmentation, as the set of genes assigned to different cells can be compared with the expression types seen in a single cell experiment with the same tissue.
It is easy to imagine even in an old-school computational framework how single cell data might be used as a prior for improving or vetting spatial cell segmentation, or for using a small number of genes resolvable with spatial to impute the expression of other genes. With modern machine learning techniques, it's ever easier to imagine this - the more there are large, consistently generated single cell and spatial datasets from the same tissue, the better such inference algorithms should perform.
Scale's platform can also be used for other single cell analysis, such as methylation profiling using bisulfite conversion. - which could be yet another input for machine learning if profiled on the same tissues.
The scientific data generation aspect is the more optimistic view of the collaboration, and I do like being an optimist. But some harsh reality breaks the rose-colored glasses. AGBT opened under the pall of nanoString filing for Chapter 11 bankruptcy protection and VizGen. 10X Genomics' aggressive patent enforcement activity forced nanoString's action, and since I chatted at AGBT VizGen has also been forced to reorganize (and scored further investment) after early setbacks in patent lawsuits with 10X. There doesn't yet appear to be 10X shots fired at Scale, but that's surprising - 10X drove 1CellBio into liquidation and has an ongoing battle against Parse Bioscience, another purveyor of split-pool technology. In such an environment, pairing up for safety makes sense - safety in numbers.
Might we see single cell and spatial companies go as far as merging? 10X Genomics is currently the only US player in both markets (BGI group has technology in both spaces). Both markets are still seeing new entrants - at AGBT there were 3-4 instrument free single cell startups with posters and my efforts to enumerate all the spatial players puts that around a dozen companies - though some are zombies. New spatial 'omics methods are still popping up regularly from academia; one showed up after I drafted this post. Is consolidation within each market or between the two markets more likely? Sales and support resources can be shared - a possible economic advantage to consolidation. Larger corporate mass can also drive better deals with suppliers, though depending on the platforms there may or may not be much shared in terms of consumables.
Even without the vigorous patent litigation, both single cell and spatial should be interesting to watch on both the business and scientific fronts. With the litigation, it will be absolutely chaotic. Heck, I'm even considering taking on reading the patents in question, if only to contemplate how hard they will be to design around. Join me on the bleachers - we might all learn something!
No comments:
Post a Comment