Catch up? Overtake???

Tasks of domestic sequencing

genseq, Geektimes

Until recently, in 2016, the undisputed leadership in the genomic sequencing market belonged to the United States. More precisely, the American company Illumina, which has developed a range of fluorescent sequencers.

Fig. 1. Fluorescent sequencers from Illumina

The improvement of these devices and the fluorescent technology used by them made it possible to reduce the cost of sequencing the human genome to $1,000 by mid-2016.

The second place in 2016 was taken by the American company Thermo Fisher Scientific, which develops semiconductor DNA sequencing technology. Their Ion S5 sequencer, despite its relatively modest performance (up to 12 Gb), quite adequately competed with Illumina desktop sequencers in the niche of targeted (clinical) sequencing.

Fig. 2. Ion S5 semiconductor sequencer

In September 2017, BGI announced the start of accepting applications for genome sequencing for $600, which immediately brought China to the leaders of genomic races. This breakthrough was made possible by the creation of CNGB (China National GeneBank), a large center with 150 fluorescent Chinese BGISEQ-500 sequencers.

Fig. 3. In the CNGB "machine room"

However, the performance of one American NovaSeq 6000 is equal to the performance of 50...60 Chinese BGISEQ-500. Therefore, the largest sequencing center today can be considered the Chinese company Novogene, which at the beginning of this year acquired 25 NovaSeq 6000 at once. Their total productivity is about a quarter of a million genomes per year. The figure is impressive, but if 0.3 million genomes are sequenced annually (~0.05 million in CNGB plus ~0.25 million in Novogene), then 100 million genomes will be sequenced under the "China Precision Medicine Initiative" program launched last year (2016...2030, $9.2 billion). it will take more than 300 years. And in order to meet the allotted deadlines (by the end of 2030), the Chinese will have to build and equip several dozen more such centers with sequencers.

At the beginning of 2017, MinION miniature nanopore sequencers appeared on sale, and in May – GridION X5, developed by Oxford Nanopore Technologies (ONT, UK). The most productive model (PrometION) is undergoing beta testing in several genomic centers and should be available in the coming months.

Fig. 4. ONT nanopore sequencers

The relatively low accuracy of nanopore sequencing (~90% with a single reading) does not allow these devices to compete with fluorescent sequencers (accuracy ~99.9% with a single reading) in determining point mutations (Single Nucleotide Polymorphisms, SNPs), but the large length of the reads (>10,000 bp) makes them indispensable for mapping polymorphisms of the type Copy Number Variations (CNVs). In addition, nanopore sequencers do a good job of identifying viruses and bacteria, assessing their drug resistance, analyzing transcriptomes, HLA typing, establishing paternity and many other tasks of targeted sequencing, which allows them to successfully compete for these niches of the NGS (Next Generation Sequencing) market.

Interference in the genomic races of China and the UK has intensified the competition. This has not yet affected the prices of targeted sequencing, but the cost of sequencing the human genome has decreased by 40% over the past year (from $1,000 to $600).

Should Russia participate in the genomic race, or is it easier to wait for the appearance of cheap Chinese, English or American sequencers? But such an expectation can be very long. And it's a shame for the state. This determines the relevance of considering the possibility of developing a domestic sequencer and providing it with consumables and reagents.

The main goal of such a development is to "catch up and overtake America" (as well as China, Great Britain, South Korea, Australia, Saudi Arabia, etc.). Or at least just catch up. Or not even catch up, but just try to make sequencing more accessible in Russia. First of all, to achieve import substitution of at least part of consumables and reagents. It will be more difficult to copy sequencers. But you can not just copy foreign developments, but try to improve them. And if not to improve, then at least to make it cheaper. The task is not too ambitious, but it is doable.

One of the projects of this kind was worked out by four institutes of the Siberian Branch of the Russian Academy of Sciences (2012...2014), which unsuccessfully tried to master the SMRT sequencing technology (Pacific Bioscience, USA). We can also mention two attempts to develop a monomolecular sequencing technology based on Raman spectroscopy – in Chernogolovka (OOO "InSpectr", 2010...2012) and in Zelenograd (OOO "Nano Vizhin", 2013...2014) – and the Zelenograd RuSeq project aimed at improving TSMs technology (Helicos, USA).

It is clear that when choosing NGS technologies to be mastered (copied/improved/"copied"), it is necessary to take into account the prospects for their development. And, given the extremely limited possibilities, to evaluate these prospects only for the three most advanced technologies – fluorescent, semiconductor and nanopore.

Fluorescent technology

In this case, the sequencers are precision scanning epifluorescence microscopes equipped with a system for supplying reagents to flow cells. A characteristic feature of the latest models is the ordering of the arrangement of submicron DNA clusters (Illumina, NovaSeq 6000) or DNA nanoballs (BGI, BGISEQ-500) in disposable flow cells.

Such microscopes in Russia will have to be assembled mainly from imported components, so they will cost no cheaper than their Chinese counterparts. True, these analogues are not sold yet, but in 2... 3 years, most likely, they will be available with us. Therefore, it is better to focus not on the development of fluorescent sequencers, but on the development of the production of their consumable components and reagents – flow cells and labeled nucleotides. Unless, of course, fluorescent technologies will not be replaced by fluorescent ones in a few years. Moreover, such a replacement may begin as early as 2018.

Luminescence has already been used in NGS – in pyrosequencing technology, which allowed 454 Life Sciences to read the first individual human genome (“Project Jim”, 2005...2007). This technology, based on bioluminescent (luciferase) registration of pyrophosphate formation, is now outdated. But it is easier to determine luminescence than fluorescence. Therefore, Illumina has been developing luminescent sequencing technology for a long time (the "Firefly" project).

The fluorescent sequencer may be no worse, but much cheaper than the fluorescent desktop sequencers MiniSeq and MiSeq, which is why its development is progressing very slowly. Nevertheless, recently at the ASHG 2017 exhibition (17...21.10.2017), a ready-made Firefly sequencer was demonstrated, as well as the flow cells (chips) and reagent cartridges necessary for its operation.

Fig. 5. Firefly Fluorescent Sequencer Illumina Companies

The main problem in the case of orientation to luminescent technology will be not so much the development of the device, as the development of the synthesis of consumable reagents necessary for its operation – deoxynucleoside triphosphates (dNTP) with tags capable of generating photons. Moreover, these labels should be connected to nucleosides by easily cleavable linkers containing azidomethyl groups.

An important feature of azidomethyl derivatives of dNTP, the synthesis of which was developed by Russian scientists (IBH RAS) in the early 90s of the last century, is their relatively high stability, combined with the simplicity and speed of unblocking when processing DNA clusters (or DNA nanoballs) with a solution of tris(2-carboxyethyl)phosphine (TCEP). It was the complexity of the synthesis of such reagents that until recently protected Illumina from competitors, and the development of their production allowed China to catch up and overtake America.

Are Russian chemists able to master the synthesis of such reagents? Judging by the references in the patents of Illumina, in the 90s of the last century there was no doubt about it. And now in Russia there are 3...4 groups of chemists capable of coping with such a task (IHBFM SB RAS, LLC "Syntol", IBH RAS, IMB RAS).

Semiconductor technology

Thermo Fisher Scientific has spent billions of dollars on the acquisition of semiconductor sequencing technology. Now the intensification of competition requires a sharp reduction in prices, and it is unlikely to be able to return the billions spent, especially with a profit. Third-party developers do not care about such problems, so for them, semiconductor technology still retains its appeal. Especially if you manage to use ready-made pH-sensor chips, for the development of which millions of dollars have been spent.

Fig. 6. pH-sensor chips of the S5 semiconductor sequencer

The cost of these chips is overstated, at least by an order of magnitude. And you can use them (according to the developers) only once. However, some craftsmen managed to use them more than ten times, and this is clearly not the limit. Therefore, the primary task for improving (reducing the cost) of semiconductor sequencing technology is to master the regeneration of used pH-sensor chips.

A prototype of the device necessary for such regeneration has already been developed. More precisely, an electronic system has been developed that reads information from pH–sensor chips and allows monitoring their quality.

Fig. 7. Homemade semiconductor sequencer

If we increase the speed and bit depth of the ADC, then such a reader can be used as an electronic subsystem of a domestic semiconductor sequencer. However, it will still need to be equipped with a reagent supply system. And master the production of these very reagents. With a strong desire (and good funding), there are no special problems with this.

The problem is that all such developments and their improvements will take 2 ... 3 years, and during this time a lot can change. For example, the accuracy and performance of nanopore sequencing may increase. As a result, all the efforts of "semiconductor" competitors will be in vain.

Nanopore technology

The first Minion nanopore sequencer is similar to the first pancake – it is already "edible", but the next ones should turn out much better. His reading accuracy is no more than satisfactory, and even then not for all applications. As for productivity, it is clearly not enough for genome sequencing, since at least five disposable cells worth from $500 to $900 (depending on their number in the order) have to be spent on each genome.

MinION cells have integrated chips that amplify and digitize signals (picoampere currents) from 512 nanopores. GridION X5 works simultaneously with five of the same cells, but in the cells to PromethION, the number of analyzed nanopores has been increased by 6 times (up to three thousand). This will allow sequencing of the human genome on a single cell. True, with low quality, but with long reeds, which facilitates their precise assembly. And it complements well the short (2x150 or 2x100), but accurate (>Q30) reads obtained by fluorescent sequencers. Therefore, nanopore sequencing on PomethION can complement fluorescent sequencing, but cannot compete with it. Although if the next generation cells contain not thousands, but tens or hundreds of thousands of nanopores, then their use will increase the multiplicity of DNA reading, improve the quality of the data obtained and allow genomic nanopore sequencers to successfully compete in the NGS market with fluorescent sequencers.

For most targeted sequencing tasks, the MinION performance (5...10 Gb) is clearly redundant. Therefore, ONT plans to launch the MinION Dx (or FLONGLE) – a modification of the MinION with an adapter insert for 128 or 256-port cells.

Fig. 8. The FLONGLE sequencer

Disposable cells for FLONGLE can be much cheaper, since their electronics are placed in a reusable adapter insert, with which they are joined by a contact pad of the LGA type (Land Greed Array).

Another compact nanopore sequencer (SmidgION), connected to a smartphone (iPhone 7...X), should be available in the coming months.

Fig. 9. The first "gadget" sequencer SmidgION

Cheap and affordable nanopore sequencing can change the entire NGS market (and at the same time the whole world). But this ability will be fully manifested only after the appearance of worthy competitors. One of such competitors may be companies Roche Sequencing, which has been developing its own nanopore sequencing technology since 2014. Judging by some publications and messages on the Internet, other competitors may soon appear.

It would be nice to acquire similar competitors in Russia, but the development of sequencers, especially nanopore sequencers, was not included in the list of "Priority areas for the development of science, technology and technology in the Russian Federation" approved by Presidential Decree No. 899 of July 7, 2011. Therefore, we can only hope for geeks who develop semiconductor sequencers in their own kitchen or electron microscopes in their personal garages. It will not be possible to do without hackers who can hack the software to MinION. The fact is that this sequencer can only work if there is an Internet connection. And each autonomous launch must be coordinated with the developers. But there are also more strange "problems". For example:

• Each device and each flow cell is linked to a specific user, who is backed by the actual address of the laboratory where he works and conducts research. At the same time, Oxford Nanopore Technologies can get information about the location of each device.

• Sanctions policy: Before the delivery of products, Oxford Nanopore Technologies checks each organization and new end users.

• In order to avoid being included in the List of laboratories Prohibited for Shipment, it is prohibited to transfer Oxford Nanopore Technologies products to third parties.

Interestingly, in accordance with the above-mentioned "List of Laboratories prohibited for shipment", ONT recently refused to sell MinION to such "paramilitary" organizations as Moscow State University and St. Petersburg State University.

In this regard, the first priority for Russia in the field of nanopore sequencing is hacking the software used by the MinION sequencer.

The next task is the reverse engineering of the cells of this device. And finally – mastering the formation of bilayer lipid membranes in them with built-in ion channels suitable for nanopore DNA sequencing. It is better to look for biohackers (biophysicists, molecular biologists, genetic engineers, etc.) for such work elsewhere, but here, on Geektimes, I would like to discuss the problems of software hacking and reverse engineering with hackers and geeks who are well-versed in electronics.

In the discussion of the article , the author , among other things , explained:
"The main patents date back to the mid-90s of the last century and now their validity periods have already ended. In addition, such technologies are not patented in Russia. They don't believe that competitors can appear here."
So as for copyrights, everything is fine here, everything can be hacked, copied, imitated, etc.

However, it is not very clear how the idea of developing domestic sequencers looks from a technical and economic point of view ... – VM.

Portal "Eternal youth" http://vechnayamolodost.ru