Biomiv
Healthspan
PAPER
Menu
Careers
Open roles
Dark Mode
Plain English Mode

All articles

November 19, 2026 · 7 min read · Cymetica Biosciences

Discoveries are hidden relationships: what an enzyme map for a rare disease shows

A new structure of the enzyme AASS has been described as a drug target for glutaric aciduria type 1. The route to that enzyme runs from a disease through a gene, a pathway, a food and a neighbouring condition, and every link was already on record. That is how discoveries are made, and it is the step Biomiv is built to help with.

On 13 November 2026 a team at the Icahn School of Medicine at Mount Sinai published the first detailed structure of a human enzyme called AASS [1]. The news coverage described it, fairly, as a map of a new drug target for a rare childhood brain disease [2] [3] [4].

We want to use it to make a different point. Look at how the researchers got to that enzyme, and you find that the discovery was not a new fact. It was a new connection between facts that were already on record. That is how most therapeutic discovery works, and it is the thing Biomiv is built to help with.

To be clear from the start: this is their work, not ours. Biomiv had no part in it. We are using it as a worked example because it shows the shape of the problem unusually well.

A disease whose own enzyme cannot be fixed

Glutaric aciduria type 1, or GA1, is a rare inherited metabolic disorder. It is caused by mutations in a gene called GCDH. The enzyme that gene makes is one step in the way the body breaks down certain amino acids, mainly lysine. When that step fails, toxic by-products build up, and they can cause irreversible brain damage, especially in early childhood [2] [3].

Care today rests on newborn screening, a diet that restricts lysine, and emergency management during illness. It helps, and preventing neurological injury is still hard [3].

The obvious target is the broken enzyme. The obvious target is also the difficult one: replacing or repairing a missing enzyme inside the brain is among the hardest things medicine attempts.

The chain that leads somewhere else

The route the field took instead runs through seven kinds of thing. Each link was known. The value is in the order.

  1. Disease. GA1 damages the brain.
  2. Gene. The cause is a faulty GCDH.
  3. Pathway. GCDH sits partway along the lysine degradation pathway.
  4. Biochemical. The harm comes from what piles up behind the blocked step.
  5. Food. The pathway starts with lysine, which comes from the diet. This is why restricting lysine helps at all.
  6. A different enzyme. AASS controls the first steps of the same pathway. Slow AASS, and less material reaches the blocked step, so less of the toxic by-product is made. This approach has a name, substrate reduction: do not repair what is broken, turn down what feeds it [3].
  7. A compound. Which means the search is now for a molecule that inhibits a working enzyme, a far more ordinary problem for drug design than restoring a missing one.

Earlier studies had tested the idea by removing AASS genetically. According to the coverage, that reduced the toxic by-products in cell and mouse models of GA1, and in a study reported in March 2026 it protected mice from neurological abnormalities even when they were fed excess lysine. It did not remove every disease-associated by-product [3].

The link that makes it safe to try

One more connection carries a great deal of weight, and it comes from a different condition altogether.

Some people are born without working AASS. Their condition is called hyperlysinemia, and it is generally regarded as relatively benign [3]; its outlook has been followed in the medical literature since the 1980s [5]. That is a fact about one rare condition. Laid next to GA1, it becomes something else: evidence that a person can live with AASS switched off, which is exactly what a drug against AASS would do.

Nobody had to run a new experiment to learn that. It was sitting in the literature of a neighbouring disease.

What the new map adds

What was missing was a way to control AASS with a drug. That is what the new paper addresses. The team worked out how to isolate the complete human enzyme, imaged it by cryo-electron microscopy, tested it with substances that speed it up or slow it down, and ran computer simulations of how its shape changes [1] [2].

The finding the authors emphasise is a flexible, hinge-like region away from the enzyme's active site. Small changes there have a large effect on the whole protein. That matters for drug design because it offers a second way in: a molecule can regulate the enzyme from a distance, known as allosteric inhibition, instead of competing for the active site [2] [3].

This is early work. It was done in the laboratory. A usable inhibitor would have to act selectively, reach the brain, and prove safe over the long term in animal and then human studies, and it would add to current care, not replace it [2] [3].

Why we call these relationships hidden

The researchers do not use that word, and we are not putting it in their mouths. It is our term, and we use it in a specific sense.

In our first post we defined a hidden relationship as one that is implicit: the pieces are published, and the connection between them is not stated where you would look for it. The GA1 story is full of them. Three kinds stand out.

  • Across kinds of thing. The path crosses from a disease to a gene, a pathway, a biochemical, a food, an enzyme and a compound. Each of those lives in a different database, a different journal and often a different profession.
  • Sideways, to a neighbour. The target is not the enzyme named in the diagnosis. It is one further up the same pathway. And the safety evidence comes from a separate condition that shares that enzyme.
  • Back to the everyday. The same chain that ends in a drug target begins with an amino acid on a plate. Diet and drug discovery are not separate subjects here. They are two ends of one pathway.

None of these links is exotic. A specialist in lysine metabolism holds all of them in mind. The difficulty is that there are thousands of rare diseases and tens of millions of papers, and no one holds all of those in mind. Somewhere in that literature are other chains of the same shape that no one has yet laid end to end.

What this means for Biomiv

This is the goal we have set for Biomiv and for BioAI, its research agent: to find hidden relationships between diseases, symptoms and indications; foods, micronutrients and plant compounds; drug compounds; genes, proteins and enzymes; pathways and biochemicals. In other words, to help with the connecting step that discoveries like this one depend on.

Here is where we are, stated plainly.

What Biomiv does today. For a condition, it ranks the micronutrients, foods, plant compounds and drug compounds whose patterns of association in the research sit closest to it, and it reads each score against a control built from shuffled data, so a score that is no better than chance is labelled as such. It maps what you type to recognised conditions and the drug compounds on record for them, and it shows new papers as they arrive.

What it does not do yet. It does not walk a chain. It will not, today, take GA1 and lead you through the pathway to AASS, or point to hyperlysinemia as the reason AASS might be a safe target. Links of one kind to another are shown side by side, not yet joined end to end. Building that traversal is the work in front of us, and the GA1 chain is the test we will hold it to.

What a lead is not

A connection that a system surfaces is a lead for someone qualified to examine. It is not a finding, and it is not advice. The GA1 work shows why: between the idea of targeting AASS and this month's structure lie years of experiments, and between the structure and a treatment lie more.

If GA1 or any metabolic condition affects you or your family, the people to talk to are your metabolic care team. Nothing here is a reason to change a diet or a treatment.

References

  1. Wu R, Khamrui S, Dodatko T, Xiang Y, Golestanian A, Pham ND, Kumar K, Shi Y, Osman R, DeVita RJ, Houten SM, Lazarus MB. Structural and biochemical insight into allosteric regulation of the human 2-aminoadipic semialdehyde synthase. Science Advances. 2026;12(46):eaej2268. doi:10.1126/sciadv.aej2268.
  2. Mount Sinai. Mapping an enzyme offers clues to treating rare genetic disease. Newsroom, 13 November 2026. mountsinai.org.
  3. Kirkin A. New drug targets for rare metabolic disease identified. Inside Precision Medicine, 13 November 2026. insideprecisionmedicine.com.
  4. Wightwick S. Enzyme structure mapped as new drug target for rare brain disorder. Drug Target Review, 16 November 2026. drugtargetreview.com.
  5. Dancis J, Hutzler J, Ampola MG, et al. The prognosis of hyperlysinemia: an interim report. American Journal of Human Genetics. 1983;35(3):438-442. PubMed 6407303. Cited by reference 1.