Quantum Annealing Speeds Up Species Tree Estimation

QUANTA, a preprint method, runs on real D-Wave quantum hardware and reports a 4100x bipartition speedup over classical wQFM.

Updated on Nov. 6, 2026 in Quantum Computing

Bold vector editorial illustration of a square quantum chip with branching gold traces splitting into a tree-like pattern in teal and mustard, evoking quantum computation applied to evolutionary biology.
Researchers posted a preprint describing QUANTA, a species-tree method that runs on real D-Wave quantum annealing hardware and reports a 4100-fold speedup over classical wQFM. AI Illustration. Upload story photo >

Researchers have introduced QUANTA, a species tree estimation method that uses quantum annealing and runs on real D-Wave quantum annealing hardware, according to a preprint posted to bioRxiv. The work is at the preprint stage and has not been peer-reviewed or shipped as a production tool.

Why it matters

Species tree estimation, which reconstructs evolutionary relationships among organisms from genomic data, has been limited by the cost of classical bipartitioning heuristics such as Fiduccia-Mattheyses and MaxCut. QUANTA is the first quantum formulation of quartet-based approaches and the first multi-locus species tree method demonstrated on real quantum hardware, marking an early step toward quantum tools in computational biology.

On the avian phylogenomic dataset, QUANTA reconstructed a species tree with a higher quartet score than wQFM, and its bipartition technique ran 4100x faster than the FM method in wQFM on bipartitions of 90-200 taxa. Its accuracy is reported as comparable to ASTRAL and wQFM under challenging evolutionary conditions, though the source does not state scaling beyond the tested range.

The players

D-Wave Systems

A quantum computing company whose quantum annealing hardware was used to run the QUANTA method.

ASTRAL

A widely used classical species tree estimation method against which QUANTA's accuracy was benchmarked.

wQFM

A classical weighted quartet-based species tree method whose bipartitioning step QUANTA accelerates and whose output it was compared against.

The details

Species tree estimation groups genes into a single evolutionary tree by repeatedly splitting sets of taxa into two parts, a step called bipartitioning. QUANTA recasts this weighted quartet bipartitioning problem as a quadratic unconstrained binary optimization (QUBO) problem, a format that quantum annealing hardware can solve by searching for a low-energy configuration. The formulation uses the minimum possible number of qubits, the quantum bits that carry the computation, to represent each bipartition.

Timeline

  1. Preprint describing QUANTA posted to bioRxiv on November 5, 2026.

The Tech Race

QUANTA sits in a race to apply quantum annealing to computational biology, following the QFM and QMC families of quartet-based methods that it reformulates. Its reported 4100x bipartition speedup over the FM technique in wQFM positions it against classical species tree tools such as ASTRAL and wQFM.

The source code is publicly available on GitHub, so researchers working on phylogenomics can inspect and test the method on their own data. Practical use depends on access to quantum annealing hardware, and the stub gives no timeline for broader availability.

The takeaway

QUANTA shows that a species tree method can run on real quantum annealing hardware, which points toward more scalable phylogenomic analysis as the hardware matures. Watch for peer review of the bioRxiv preprint and for independent benchmarks on taxa counts beyond the 90-200 range reported.

Further reading

For background on the hardware behind such experiments, see the Quantum Computing section.

More information

Download the QUANTA publicly available source code repository to test the method directly.

Source note: This article includes information reported by Biorxiv.