A cookbook without a shopping list: that's what a scientist faces when looking at a scheme of chemical reactions. The arrows and formulas on paper look neat, but once you start hunting in the lab for molecules that will actually behave that way, a real quest begins.
The authors based their work on a strict rule: in every reaction, the number of atoms on the left must equal the number on the right (this is called conservation of mass). Seems like simple arithmetic. But when you need to pick specific substances from a huge catalog of known molecules, while also staying within allowed types of chemical transformations, the task turns into a decision-making puzzle of incredible complexity.
Yet a way out was found. The researchers used a backtracking algorithm with pruning: it's like playing chess blindfolded: you try moves, but ruthlessly discard those that lead to dead ends. This approach, inspired by searching for identical fragments in graphs (the subgraph isomorphism problem), makes it possible in practice to assemble working molecular property prediction schemes for stochastic modeling of complex systems — from ecology to medicine.
The recipe book takes on flesh. And behind this is not magic, but precise calculation that teaches us: even in the chaos of chemical possibilities, there is a strict logic.
Many biochemical networks in the cell still cannot be reproduced 'in a test tube' precisely because we cannot pick the right participant molecules — even though all the reactions are known on paper.