“Here we have to run as hard as we can just to stay in the same place. If you want to get somewhere else, you have to run twice as fast.” The line is from Lewis Carroll’s Alice’s Adventures in Wonderland: Alice and the Red Queen set off together, only to find themselves back at the same starting point.

Biologist Leigh Van Valen proposed the Red Queen theory from this idea in 1973. In essence, biological systems achieve evolutionary innovation in long-term dynamic equilibrium. For example, viruses keep evolving stronger infectivity while hosts evolve stronger immunity; both keep innovating and forming new balances.

Understanding coevolution from structure: compute demand rises

The COVID-19 coronavirus outbreak began in late 2019 and swept the world in early 2020. Before that, six other coronaviruses had infected a limited set of hosts, including humans and birds; four of them were already circulating in the human population.

Over billions of years of evolution, viruses and hosts have coevolved in the Red Queen manner. The time and place of this outbreak are still not exact, and the intermediate host is still unclear.

That Red Queen interaction and coevolution also pose a great research challenge and have driven new algorithms that understand coevolution from a structural angle. Analyzing large volumes of genomic data, identifying key sites under positive selection, and running dynamics simulations all consume large amounts of compute.

Sugon computing: seeking coevolutionary signals

Studying coordinated-evolution signals in virus–host interaction genes is important for understanding evolutionary history, finding intermediate hosts, and preventing coronavirus outbreaks. With computing support from Sugon, a genomics-center team will use molecular-dynamics simulation to predict how different SARS-CoV-2 strains infect different human populations.

Drawing on the Red Queen theory, the team proposed a new method to explore coordinated evolution between virus and host. By analyzing positively selected sites in virus–host interaction genes and using homology modeling, a relatively reliable protein-structure method, they run molecular-dynamics simulations on the viral S-protein receptor-binding domain (RBD)–host ACE2 complex to predict binding energy and explore how, and how tightly, they bind.

In short, this approach offers a feasible path to study virus–host coevolution and to explore potential intermediate hosts.

Computing and life-science research growing together

In the life sciences, protein analysis, gene sequencing, molecular dynamics, computer-aided drug design, and medical imaging all demand large amounts of compute. Those demands have also pushed the computing industry and its service models to mature.

In July, Sugon launched Computing Services, a compute-infrastructure and public computing platform. By linking data-center computing resources nationwide, it interconnects capacity, schedules it uniformly, and provides on-demand, metered, one-stop services covering compute, applications, data, tuning, consulting, operations, and maintenance—offering life sciences and other compute-hungry fields massive capacity that is as ready as turning on a tap.

Sugon Computing Services can provide many-core compute nodes for fast single-core multithreaded programs, and SSD storage for stable high-IOPS workloads. It can also offer data-sharing services for bioinformatics companies and APIs that integrate into a company’s own platform.