SCIENTISTS are suckers for stories of discovery. In 1953, James Watson and Francis Crick went for a drink in The Eagle pub in Cambridge after having worked out their model of the DNA double helix.

According to Watson, Crick announced to the other bemused drinkers: “We have found the secret of life.”

A great story — except that, according to Crick, he never said it. But the story became so memorable that it stuck.

Fifty years later, Watson unveiled a blue plaque commemorating the event at the pub, with “the secret of life” at the top.

By then he was well-known within the scientific community for his “flamboyantly chauvinist” views.

His misogynistic portrayal of Rosalind Franklin in his 1968 book the Double Helix had become a legendary example of scientific sexism.

And by the time he unveiled the plaque he was also known for his claims that racist stereotypes had some genetic basis. It must have been comforting to dwell instead on his past glory.

In 2016, Watson admitted he had made the phrase up “for dramatic effect.” But the myth caught the public imagination.

Solving the structure of DNA could be called “the secret of life” because the double helix showed how the sequence of a chemical could store information.

DNA is a double helix. Imagine this as two vertical parallel strands winding around each other, with connections across between them: a twisted spiralling ladder.

The rungs of the ladder are made up of four chemical constituents: adenine (A), cytosine (C), guanine (G) and thymine (T). (The letters are just for simplicity.)

These “bases” of DNA are such that each type will only bind to one of the others: A with T and C with G. Every “rung” is made up of one of these “complementary” base pairings.

Going along one strand of the ladder, the sequence of the bases on that side — for example, AGGC — will always be “mirrored” on the other strand.

Imagine going backwards through AGGC that sequence starting from the end: C binds with G, G with C, another G with C and A with T. So, the other strand will read GCCT.

The sequence could somehow store information that could be translated into proteins, which can interact in the cell to produce life’s complexity.

The double helix means DNA is both a secure store for genetic information and also has an obvious means of copying the information.

If you were to unzip the helix, breaking each rung in two, you’d have two independent strands.

Any one contains exactly the same information. If you add in a soup of many base molecules and the right conditions, you can make each strand into an identical double helix as the bases snap into place along the broken ladders.

But DNA was only part of the puzzle. Once proteins were made, how did they interact with each other in the cell? This was a problem studied by the French scientist, communist and one-time member of the French Resistance, Jacques Monod.

In 1961, Monod (supposedly) walked into the laboratory of his colleague, Agnes Ullmann.

He looked tired and worried and after a few minutes of silence said: “I think I have discovered the second secret of life.” After a few glasses of whisky he explained.

Proteins are made as a long chain of amino acids, but then fold up into complicated three-dimensional shapes.

Compared to DNA, they are much more flexible and active. If DNA has evolved to be as inert as possible, in comparison proteins have a can-do attitude.

A protein usually functions by binding to something else. Many molecular pathways in the cell involve an exquisite sequence of proteins binding to each other.

This had been thought of as a lock and key mechanism: like a lock, each protein could perform one job, but do it perfectly.

Monod’s realisation was that protein structure offered a way for the same protein to change what it did. Say a protein called X binds some other molecule Y at an “active site” somewhere on its structure.

It would be possible that somewhere else on the structure of X, X could bind with another molecule Z.

This binding could be such that this could change the shape of X’s active site with Y — perhaps such that Y could no longer bind.

This might seem abstract, but it is in fact very practical: changing the amounts of Z in the cell will change whether X binds Y or not.

This ability to change structure means that proteins are not static locks which can only ever fit one key: the lock is dynamic.

Their function can be “switched” by other molecules. This transmission of binding at one site to another on the same protein is called “allostery.”

Fundamental to life is the possibility of changing what you do in response to conditions. Allostery is an important way to do this.

Allostery is over 50 years old as a concept. But researchers still usually only study the protein’s active site directly. Proteins are made of amino acids, of which there are 20 possibilities. Mutations cause changes in proteins.

A different one at a single place could produce allosteric effects. Many disease-causing mutations in humans are known to be allosteric, but investigating them is challenging.

Predicting known protein structures is one thing — now thought to be largely “solved” by machine learning — but predicting the effect of a new mutation on the overall structure remains extremely hard. More experimental data is needed.

The issue is that a mutation on a protein can alter the rate at which it binds something else in two ways. One way is more boring: it could change the stability of the protein, meaning that it doesn’t fold up into its usual structure.

More interestingly, it could allow the protein to fold properly, but change the active site through allosteric effects. To disentangle allosteric interactions, one needs to measure both possibilities.

Last month, researchers based in Spain published an article in Nature outlining a method that does this. With a set of relatively simple experimental tests, their method quantifies both the abundance of the folded protein and its binding to another protein.

By generating all possible mutated versions of the protein and running the same measurements on each, they can generate an allosteric “atlas” for a protein: a measure of what changing different sites does.

They studied two well-studied protein-protein interactions from humans. What they found was that around half of all places on the surface of the folded protein structures — far away from the active site — had at least one mutated possibility that was allosteric.

This result suggests that allosteric interactions are frequent in proteins. Future research generating all the possibilities for a protein might be possible, giving us more insight into not only how proteins can evolve but how we might design better drugs to manipulate them.

The researchers’ paper doesn’t have a neat discovery story. As is often the case in science, there is no eureka moment — instead, hundreds of thousands of experiments and patience produce a fascinating new understanding of an old concept.