At the Weizmann Institute of Science, Aviv Shoshany placed millions of engineered viral variants into a petri dish and watched evolution unfold in real time—compressing three years of pandemic change into just a few months. What emerged was a laboratory recreation of one of humanity's most consequential biological stories: the evolutionary path from the original COVID-19 Wuhan strain to the highly contagious omicron variants that came to dominate global infection patterns by late 2021.
The achievement marks a remarkable convergence of scientific foresight and hard-won insight. In August 2021, Prof. Gideon Schreiber published findings from an earlier in vitro evolution experiment identifying a pair of mutations in the coronavirus's binding site that would make the virus exceptionally good at attaching to human respiratory receptors. Three months later, when the omicron variant was first identified in South Africa and sequenced, researchers found the exact same pair of mutations. That moment—seeing the lab's prediction validated in real-world virus—prompted Schreiber to ask a larger question: Could this method reveal not just what mutations might emerge, but how pandemics themselves evolve?
In a new study published in Nature Communications, Schreiber's laboratory at the Weizmann Institute collaborated with Dr. Jiří Zahradník's group at Charles University in Prague to test this possibility. The method was deceptively elegant. The researchers replicated the gene encoding the coronavirus binding site using a deliberately error-prone mechanism, simulating mutation in "fast forward." They then exposed millions of resulting variants to human receptors, keeping only those that still bound successfully—mimicking natural selection at scale. By repeating this cycle of mutation and selection again and again, the team reconstructed the entire arc of viral evolution that unfolded across billions of human infections.
The critical finding emerged when the researchers tested two different evolutionary scenarios. Under strong selection pressure—where only a small number of viruses survive each stage, allowing advantageous mutations to dominate rapidly—something striking happened. "No matter which viral variant we started with, under strong selection pressure a variant remarkably similar to omicron and its sub-variants emerged early on and rapidly took over the entire population," Schreiber said. This trajectory matched precisely what the world actually witnessed during the pandemic, which has not undergone another major shift since omicron became dominant at the end of 2021.
Under weak selection pressure, by contrast, where many viral variants persist simultaneously, the evolution proceeded differently. This scenario was simulated by researchers in Czechia and revealed how environmental conditions shape viral evolution in fundamentally different ways.
The implications reach far beyond COVID-19. Schreiber notes that "some future pandemics that spill over from animals to humans may follow a similar path—accelerated evolution culminating in the dominance of a viral variant that is highly contagious and specifically adapted to bind to human receptors." When the team subjected the original SARS virus from the early 2000s to the same experimental conditions, variants that bound very efficiently to respiratory tract receptors emerged rapidly, suggesting this evolutionary pattern may be common across coronaviruses.
The work offers something science rarely delivers: the possibility of prediction. By understanding the conditions under which viruses evolve toward higher contagion, researchers may eventually forecast which variants are likely to emerge and when—turning pandemic response from reactive to anticipatory. For now, the test tube has shown us that evolution, given certain conditions, follows surprisingly predictable paths.