Background
The Board of Trustees of Leland Stanford Junior University sought a patent for a computational method for haplotype phasing — a bioinformatics technique for determining which genetic variants (alleles) are located together on each copy of a chromosome. Human chromosomes come in pairs, and knowing which genetic variants are inherited together (i.e., on the same chromosome) versus distributed across the two copies is important for understanding inherited diseases and population genetics. The claimed method performs this determination using statistical calculations run on a computer.
The USPTO rejected the application under 35 U.S.C. § 101, concluding that the claims were directed to abstract mathematical algorithms. Stanford appealed to the Patent Trial and Appeal Board, which affirmed, and then to the Federal Circuit.
The Court’s Holding
The Federal Circuit affirmed, holding that the claims were directed to an abstract idea — specifically, a mathematical algorithm for haplotype phasing — and that the implementation of those calculations on generic computer hardware did not supply an inventive concept sufficient to make the claims patent eligible.
Under Alice step one, the court found that the core of the claimed invention was a series of mathematical calculations and probability computations used to infer haplotype structure from genetic data. Mathematical algorithms, however useful or novel, fall within the abstract idea exception to patent eligibility. The fact that the algorithm served a valuable purpose in genomics research did not change its character as a mathematical method.
Under Alice step two, the court found that the claims added nothing to the abstract algorithm beyond the use of generic computer hardware — a processor running the calculations. Using a general-purpose computer to perform mathematical calculations has never been sufficient to transform those calculations into patent-eligible subject matter, and Stanford’s claims did not recite any novel computer architecture, any improvement to how the computation was performed at the hardware or system level, or any other technical improvement beyond the algorithm itself.
Key Takeaways
- Mathematical algorithms — even novel, scientifically valuable ones for applications like genomics — are abstract ideas under Alice step one, regardless of how complex or useful the math is.
- Implementing a mathematical algorithm on a generic computer does not make the claims patent eligible; the computer must be used in a way that reflects a specific, concrete technical improvement to the computer or the computational process.
- Bioinformatics and computational biology patents face heightened § 101 scrutiny when the core claimed innovation is a new statistical or mathematical technique applied to biological data using conventional computing tools.
- Academic institutions seeking to patent computational methods in genomics, biology, or related fields should focus on claims that recite specific technical implementations with concrete system-level improvements, rather than claims centered on the mathematical technique itself.
Why It Matters
Bioinformatics — the use of computational methods to analyze biological data — is one of the fastest-growing fields in science, and patent eligibility for computational biology innovations is an important ongoing question. The Stanford haplotype phasing case reflects the Federal Circuit’s consistent application of Alice’s limits to mathematical algorithms in scientific applications: the novelty of the mathematical technique, and the value of its scientific application, do not by themselves confer patent eligibility.
For universities, research institutions, and biotech companies seeking to patent computational methods in genomics and related fields, the ruling underscores the importance of drafting claims that emphasize the specific technical means by which the algorithm is implemented, not just what the algorithm does. Claims that focus on a novel way of organizing, processing, or analyzing data at the computational system level — rather than simply claiming the mathematical method — are more likely to survive § 101 scrutiny in the bioinformatics space.