Further thoughts on AxBio and Roswell
I recently wrote about AxBio and Roswell Biotechnologies. When looking into a company, I try and get some sense of existing work that might support the approach they’re proposing.
What both AxBio and Roswell are attempting is clearly very hard. But in this post, I thought it would be interesting to compare these two companies in terms of how well supported the approaches are.
Roswell
In Roswell’s case, what they’re trying to do is fairly ambitious. But there is at least literature that supports the basic detection approach. A number of groups have imaged single DNA strands on surfaces using tunneling microscopy:
So, while it’s not clear that you can extend this to detecting differences between nucleotides (and many have tried), or that you can obtain the required resolution at speed, it’s clear that there is at least some support for the basic sensing mechanism.
By starting from the basis of a known sensing method, they at least have options in terms of iterating over the approach, or pivoting to detecting a different analyte.
This seems to be what they are doing, shifting away from DNA to larger molecules that are easier to detect, and don’t require the same readout speed. This is a nice option to have, and something that’s only available to you if you have a solid foundation in a sensing approach that has been proven out.
AxBio
With AxBio there isn’t as much to go on. There are no examples in the patents, or other disclosures which would clarify exactly which approach they’re working on. So we’re mostly left guessing. I focused impedance sensing, at least in part because it’s one of the more novel ideas in their patents.
The problem is that there’s no literature supporting detecting DNA through impedance measurements. This puts us in a tough position. The fact that there is no literature already raises concerns, but we would prefer to better articulate the difficulties involved.
In this case, I looked further to afield to scanning capacitance microscopy (SCM). This gives some sense of the resolutions achievable using this sensing mechanism. What we find is that SCM resolutions are nearly an order of magnitude lower than possible with tunneling. And it feels like there would need to be some fairly fundamental scientific developments to improve this.
Conclusion
With Roswell we have a basic sensing approach, supported by the literature. With AxBio the basic sensing approach doesn’t have support in the literature. So we’re left with the sense that AxBio’s impedance approach is at least an order of magnitude harder than what Roswell are attempting. But of course, there’s a slim chance they might be able to perfect a fundamentally new sensing approach, leap frogging the competition.
I’ll be watching both companies with interest.