Twinstrand Biosciences

This post originally appeared on my substack newsletter.


Twinstrand is a University of Washington spinout based in Seattle. Crunchbase lists Twinstrand as being founded in 2015 which is not long after the foundational work was done (in 2012), they’ve recently raised a series B of $50M bringing their total to $73.2M. I see 62 employees on LinkedIn.

Approach and Applications

The basic play is that Illumina sequencing has an error rate that’s too high for some applications. To me, this is was kind of surprising. In Illumina sequencing, around 90% of bases are Q30. That’s an error rate of 1 in 1000. Do you really need an error rate lower than this? Twinstrand propose a number of applications, these are largely around very low level mutations.

  • Detecting residual acute myeloid leukemia (AML) after treatment.
  • Mutagenesis assays, for chemical and drug safety testing.
  • Cellular Immunotherapy Monitoring

In general, I’m used to seeing plays (like GRAIL) around cancer screening. But this is aimed more at cancer monitoring. The US national cost of cancer care is $150B, there are around 1.8M cancer cases. So, if we assume that this test will be required for cancer monitoring of every patient, and yields $1000 in profit that’s $1.8B in profit. Probably enough to support the company, and make investors happy…

But for the Twinstrand play to work, and justify a healthy valuation, at least the following needs to be true:

  1. “ultra-high accuracy” is needed for cancer monitoring.
  2. The Twinstrand approach is a practical method of generating “ultra-high accuracy” reads.
  3. The Twinstrand approach is the only and best way to get “ultra-high accuracy”.

The first may be true, but it’s obviously not what GRAIL and other players have been working on for early stage cancer screening, where the focus has shifted toward base modification/methylation.

As to Twinstrand’s practicality? Hopefully we can gain some insight by reviewing the approach.


The technique relies on adding two pieces of information to double stranded DNA. The first is a unique index (a UMI) which uniquely identifies each double stranded fragment. The second is a strand-defining element (SDE). This a marker that allows the two strands forming a double stranded fragment to be distinguished.

Twinstrand use two UMIs. One of each end of the original double stranded fragment. They call these two UMIs “α” and “β” in the figure below. 

The Y shaped adapters (labelled Arm 1,2) in the diagram above introduce an asymmetry between the strands. This provides the strand-defining element (SDE) described above.

To make this clearer I decided to break to the diagram further, showing the individual amplification steps involved:

Post amplification, and in 5’ orientation you will get 4 distinct read types as shown above. Each of these can be classified as coming from either the forward or reverse strand of the original dsDNA fragment.

From there it’s obvious that you can use this information to filter out errors that occurred during amplification (including bridge amplification):

For amplification errors to propagate they’d need to occur at the same position, and of the same base. So, I’d assume a ballpark estimate is somewhere around Q60… and their reports include identifying mutation frequency down to a rate of 10^-5.


Wow, great! Q60 reads, who wouldn’t want that!

Well the major problem is that you’re going to throw away a lot of throughput. At best you will need to sequence each strand 2 to 4 times. This might be fine if you have an amplification step in your protocol anyway. Much like UMIs the Twinstrand process will just provide additional information removing error and bias.

But unlike UMIs you want to optimize for duplicates. And not just duplicates but duplicating starting material a fixed number of times. I.e. the ideal is probably to see ~4 different sequences for every original fragment of dsDNA (one of each type).

In practice, this is problematic, in their patent they state “3.1% of the tags had a matching partner present in the library, resulting in 2.9 million nucleotides of sequence data”. As far as I can tell the input datasets was 390Mb of sequence data. Processed, corrected reads therefore represent about 0.75% of the input dataset. This is a huge hit of your throughput.

The above describes the original IP, from ~2012. Most of their patents appear to be based around this basic process. However a patent from 2018 looks like it might be worth digging into in more detail. In this patent it looks like they try to more closely model errors that occur during the sequencing process (incorporating fluorescence intensity information into a two pass basecalling process).

Dreampore – Protein Sizing

This post was originally published on substack.

I came across another patent from the Dreampore folks when investigating the platform, and thought it might be interesting to write up. It’s actually not clear if this patent has been licensed to Dreampore, but 3 of the inventors are listed as working at Dreampore so it seems reasonably likely.

The patent appears to be based around work from a 2018 paper. However the patent itself is interesting to me in part because of the way it’s written and the scope of the claims.

The patent is framed around identifying impurities (protein fragments) in a 96% pure sample. They state that “it is impossible to identify these fragments by classical mass spectrometry or by HPLC”. As in their publications they use aerolysin nanopores, a three chamber device is presented:

It’s not clear to me why you want to use three chambers. They suggest that pure product could be removed from the second chamber… but don’t describe how or why this is useful. They also don’t appear to use a three chamber device in the paper.

They show what appears to be experimental data from translocations of a protein (RR10). While it’s not stated in the patent, it seems clear from the publication this is a 10 amino acid long arginine homopeptide (RRRRRRRRRR). What they shown is that they and distinguish between homopeptides of lengths between 10 and 5 amino acids.

The paper further clarifies this showing samples containing single peptide types, as compared to the mixture:

They then take this further, and show relative concentrations of peptides in a sample from a supplier claiming 98% purity:

So, overall they show size determination at a the single amino acid level and present an application in the classification of impurities during reagent production. As far as this goes, this is fine.

The patent then describes how this might be used to measure enzymatic activity. The idea is to quantify enzymatic activity on the single cell level. Enzymes are captured on a solid support. You then flow in the enzymes substrate and pass the product through a nanopore. This seems like an interesting research problem, but I’m less clear on the market for such a device, and there’s no experimental data. 

The patent overall seems quite weak (I don’t blame the authors for this, it seems like it might have been rushed).

The single independent claim in this patent reads:

“The use of an aerolysin nanopore or a nanotube for the electrical detection of peptides, protein separated by at least one amino acid and other macromolecules such as polysaccharides or synthetic or natural polymers present in a preparation where said nanopore or nanotube is inserted into a lipid membrane which is subjected to a difference in potential of over −160 mV, in a reaction medium comprising an alkali metal halide electrolyte solution with a concentration of less than 6M and at a temperature of less than 40° C., and where said use is intended to differentiate said peptides, proteins and other molecules according to their length and their mass.”

Which seems very narrow. And what does over -160mV mean? If I look to the paper I can understand that they term -100mV as greater than -50mV, but it’s obviously not in terms of absolute magnitude. And shouldn’t this be specified in relation of the cis/trans side of the pore? The specification doesn’t provide much clarity here.

There are numerous errors in the patent for example confusing voltage and current (“voltage of 25pA”). And other statements and typos that make me think the patent was rushed. Which seems like a shame…

In terms of the technical approach itself, I’m not sure if measuring reagent impurities is a compelling market. But being able to detect single amino acid differences seems like a step toward some more interesting applications.

Dreampore – Nanopore Protein Sequencing

This post was originally published on substack.


Dreampore is a French Nanopore Protein sequencing company based in Paris.

There’s not much information available on Dreampore, most of it comes from a Genomeweb review from December 2019. In this article they state that Dreampore has raised €600,000, and they have four employees. As far as I can tell from their company registration they were founded in 2018 and currently have 3 to 5 employees. According to LinkedIn, the CEO (Luc Lenglet) is also leading two other companies. I could only see one current employee on LinkedIn who appeared to be fully dedicated to the company.


Surprisingly I wasn’t able to find a patent covering the work presented in their Nature paper on protein sequencing. So this review is based on the publication only.

The work uses a protein nanopore, and detects molecules as they pass through and block a bias current. This is much like other forms of nanopore DNA sequencing. 

The publication builds on a previous work where they detect translocations of >7mer arginine (R) homopeptides. I’ll be covering this in a future post, because it’s kind of interesting in its own right. But essentially these RRRRRRR peptides block the aerolysin pore for a detectable duration. In the sequencing paper they use xRRRRRRR peptides where the x position varies. The poly-arginine region helps the peptide stick around in the pore long enough to be detected. But the idea is that the blockage current varies enough based on the single differing position.

And histograms suggest that in most cases it does:

When you look at the full set of amino acids, current blockages are less well separated:

The plots above use Ib/I0. This appears to be the signal normalized against the baseline current. It’s not super common to do this, and I wonder why have normalize against the baseline, rather than just measuring the offset against the baseline in pA. Possibly their measurements vary significantly with buffer concentration…

The raw ABF files (which suggests measurements were taken on a Axopatch) are available. So it’s possible to confirm this. But the scaling makes the plots a little harder to interpret. From example traces it looks like blockages are probably between 60 and 70pA (they all appear to be 0.3 and 0.4 in scaled units, and a typical baseline current appears to be ~100pA). So, you’re cramming 20 states into ~10pA. The best you’re like to do in terms of noise is likely ~1pA RMS at 10KHz.

They have a plot in the supplementary information which shows that in practice, they get about 10 pA of peak-to-peak noise on blockages.

From the supplementary information the average dwell time seems to be ~5ms (which remember is for 8 amino acids). So, let’s say 1ms per AA. So if we average down to 1KHz, we can probably get this to ~1pA of noise. 

It seems likely that if they attempted sequencing, multiple positions are likely contributing to the signal. Let’s be conservative and say 3 positions. For 20 AAs that means 8000 possible combinations. So I’d speculate this comes down to:

0.001pA difference between each state and 1pA of noise

Which seems like a very hard problem to solve. Certainly one or two orders of magnitude harder than nanopore DNA sequencing.


The positive side of this paper, is that they’ve clearly shown differences between most amino acids. In practice, I don’t think these differences are good enough to clearly differentiate between all 20 AAs. But it does indicate that if you had a way of sufficiently slowing the translocation of a protein you might be able to show some kind of characteristic signal.

The remaining problems are however two fold:

  • How do you slow the translocation of proteins sufficiently.
  • How to you deal with contributions from adjacent bases. 

Both these problems are pretty tough. On the plus side, we likely only need to generate a characteristic fingerprint for a protein to be able to address useful applications. But even to get to that point, the above problems likely need to be addressed.

This paper suggests that with further work, it might just be possible. I’ll be keeping an eye on this and other nanopore protein sequencing approaches, as any kind of usable data from such a platform would be pretty exciting.

SBIR – America’s Training Program?

I’ve been thinking about the US SBIR (Small Business Innovation Research) grant program. The SBIR program gives grants, generally to what would be called “deep tech” companies. The grants are supposed to fund research and product development. I’m mostly familiar with the genome sequencing technology grants (which come via the NHGRI). These are grants on the order of ~$200K to a few million.

The SBIR program is also pretty unique. Elsewhere, government grant funding usually takes the form of matched funding. To a seed stage startup, matched funding is largely useless… after all if you’ve got no money what are you going to “match” it with. So, US SBIRs are unique in this regard.

SBIRs always look appealing to someone looking at the US funding ecosystem from the outside. Illumina received an SBIR in 1999 when it was only a year old. But at the same time, while probably $10M+ has been given out to DNA sequencing companies via SBIRs, the dominating technology was developed in the UK and acquired by Illumina. So if the purpose of these grants was to promote technological innovation… it doesn’t really seem to have worked.

Recently I’ve been thinking about SBIRs in a different light. Perhaps it’s better to think about the SBIR program as America’s technology training program. Illumina could have built out research and development pretty much anywhere. After the Solexa acquisition they could have built out operations in the UK. And to an extent they did. But the bulk of Illumina’s growth has occurred in the US.

Some of this is because they’re a US company and that’s what US companies do… but I’d also suggest that it’s just easier to hire scientific staff in the US. It’s easier in part, because there’s a community of (lets face it often failing) SBIR funded research companies, which have trained up staff in skillsets useful to companies like Illumina.

I decided to test this hypothesis with a quick LinkedIn search. I looked at 20 employees at Illumina who work in a scientific roles, in the US. Of these 50% had previously worked at SBIR funded companies, often quite early in their careers.

It was likely this previous scientific experience that made them attractive to Illumina. And is no doubt one of the reasons that Illumina would find it easier to build out research and operations in the US.

So.. the SBIR program while might in some respects look like a failure… it probably does a lot of help the US maintain technological dominance in certain industries.