What scientists miss when they don’t read patents
Why my lab reads patents to understand plant immunity, and the wiki we built so you can too.
A tomato immune receptor used in dozens of commercial varieties appears in an Enza Zaden patent as SEQ ID №116. In the Korean member of the family, Google’s machine-translated title promises tomatoes resistant to “Tomato Brown Lugos Fruit Virus,” a misspelling of Tomato brown rugose fruit virus (ToBRFV). A useful reminder that this literature can take a little deciphering.
We know the receptor as HREZ. More on that below.
When my lab looks for interesting plant immune receptors, we read papers. We read preprints. We also read patents.
The literature nobody assigned you to read
Patents are hardly everyone’s idea of recreational reading. The prose can be obscure, repetitive, the gene names elusive, and the data and sequences buried in documents that were clearly not designed for a journal club. But there is biology in there. Sometimes, biology that has yet to attract much attention from academic researchers.
Disclosure is part of the patent bargain. Published patent documents can contain detailed experiments and sequences. That information deserves to be read and assessed by scientists.
Plants use intracellular immune receptors called NLRs to detect pathogens and activate defence. Some of these receptors protect crops against devastating diseases. A receptor identified to solve an agricultural problem can help us discover a biological principle.
The catch is that patents rarely put the biology where you would look for it. In the current NLR patent wiki collection, 145 of 218 NLR or NLR-related families mention NLR terminology only in the description, rather than the title, abstract or claims. HREZ is one of them: its patent describes the encoded protein as an NBS-LRR protein, an older term for an NLR.
An old habit
We were already mining patents when Jiorgos Kourelis — now an Assistant Professor at Imperial College London — compiled RefPlantNLR, our reference collection of experimentally validated plant NLRs. Jiorgos, Toshiyuki Sakai, Hiroaki Adachi and I published it in 2021. That version contained 481 receptors. When sequences were unavailable in GenBank, Jiorgos retrieved them from other sources, including patents.
We were stunned at how much interesting science was buried in patent databases yet rarely reached the formal scientific literature. Restricting ourselves to journal articles would have left useful information behind.
Once you start looking this way, patents become another part of the scientific literature, with evidence to assess and leads to follow.
From a seed company patent to root immunity
Consider MeR1, an NLR that confers resistance to the root-knot nematode Meloidogyne enterolobii. Researchers at Rijk Zwaan, the Dutch vegetable seed company, described it in a patent published in 2021. They also reported that MeR1 co-segregates with NRC6, one of the helper NLRs we study.
MeR1 subsequently became part of our investigation of immune receptor networks in tomato. Daniel Lüdke and colleagues showed that MeR1 signals through the helper NLR NRC6 to trigger cell death. MeR1, the cyst nematode resistance gene Hero, and NRC6 belong to a specialised sub-network of immune genes that are expressed almost exclusively in roots.
That finding opened up a broader question: how does plant immunity specialise between organs? A root and a leaf encounter different parasites. How has the immune system evolved to accommodate that difference?
The agricultural problem and the biological question belonged in the same investigation.
Back to SEQ ID No 116
Back to HREZ, the tomato NLR that confers resistance to Tomato brown rugose fruit virus, or ToBRFV. This virus can overcome the widely deployed resistance provided by Tm-²².
HREZ practical relevance is unusually tangible. In their published account of the breeding programme, Enza Zaden researchers report introducing 18 commercial tomato hybrids carrying the resistance in spring 2022. By August 2025, their portfolio included 67 HREZ varieties.
Yet a receptor can be useful in crops while its activation mechanism remains poorly understood.
Get KamounLab’s stories in your inbox
Join Medium for free to get updates from this writer.
In their October 2026 preprint, Yu Sugihara and colleagues show that HREZ responds to the viral movement protein and assembles into a higher-order complex upon activation. AlphaFold 3 modelling supports a pentameric resistosome model; the precise architecture remains a prediction. Their comparisons with Tm-²² indicate that distinct NLR lineages converged on recognition of tobamovirus movement proteins, while retaining different recognition spectra.
Quite a journey for SEQ ID №116, from a Korean patent published in 2020 to a resistosome model in 2026. There is still plenty to investigate about how recognition evolves, how receptors activate, and what determines the range of viruses they detect.
HREZ and Tm-²² belong to distinct NLR lineages, yet have convergently evolved to recognise tobamovirus movement proteins. That raises a question: how many other NLRs have independently evolved to detect these proteins? Ten? A hundred? A thousand? What’s your guess? Now that’s a BBQ — a Big Biological Question, for the uninitiated.
Gained in translation
Solving a practical problem can expose a whole set of mechanistic questions. I have written about this as “gained in translation”. Projects with applied goals can enrich our basic understanding of the system under study.
This is one reason I find NLRs described in patents attractive research subjects. They come with a connection to an intended use, often resistance to a pathogen that matters to growers. Someone has already identified a problem worth addressing and a gene that might help. Our job is to ask how it works.
I have long been uncomfortable with the way we divide research into “basic” and “applied”, as though the starting motivation determines the kind of knowledge that can emerge.
Curiosity can lead to an application nobody anticipated. Equally, an effort to solve a specific problem can reveal something unexpected about how biology works. We should give researchers room to follow those surprises.
The question I want to ask is simple: what can we learn from this gene?
So we built a wiki
To make this literature easier to explore, we built the NLR patent wiki with Cloclo (aka Claude), our AI assistant.
The version dated 9 October 2026 organises 472 patent families returned by its search and seed collection. Of these, 218 are classified as NLR or NLR-related. You can browse by crop, pathogen, gene class and technology, then follow links to the original documents. The full table is downloadable.
A few things surprised me. If you think NLR patents are a corporate sport, think again: universities and institutes are assignees on 155 of the 218 NLR families, companies on 66. Chinese assignees appear on 125 of them. And the search returned only 7 of the 13 seed patents in its window. The rest describe resistance genes without ever using NLR vocabulary, which sums up the whole problem rather neatly.
It is an evolving, AI-assisted finding aid, with explicit limits on coverage and classification. The AI classified each entry from the patent text without line-by-line expert review; 46 records are flagged for checking. Its purpose is to help researchers locate disclosures and inspect their evidence.
Free to read and open to improve
The wiki is free and carries no advertising. If it saves you time, consider a donation to GetGenome, the charity that expands access to genome sequencing and training, particularly in the Global South.
I hope it encourages more colleagues to explore this oft ignored part of the scientific literature. There are resistance genes worth investigating and questions still waiting for experiments.
Follow the biology. Sometimes it comes with a patent number.
Acknowledgements
I’m grateful to Jiorgos Kourelis for first pointing out the value of looking at patents, and several other colleagues in academia and industry for their inspiration. This article was drafted and edited with assistance from Claude Cowork and ChatGPT. The views, interpretations and any remaining errors are my own.
Disclosure: I have filed patents on NLR biology, collaborate with the industry, and co-founded companies working on plant disease resistance. I am a co-founder of GetGenome.
Cite as: Kamoun, S. (2026). What scientists miss when they don’t read patents. Zenodo. https://doi.org/10.5281/zenodo.23541205
