Stanford University applied for a U.S. patent for statistical methods of predicting haplotype phase. In 2019, the Patent Trial and Appeal Board rejected the application as ineligible subject matter. Last week, a panel of the U.S. Court of Appeals for the Federal Circuit affirmed.
The opinion is interesting for being related to two, often separate fields: software and medicine (specifically genetics and bioinformatics). So, let’s look at what the Federal Circuit said about this computerized method for haplotype estimation.
Gene sequencing is common for determining the genes present in an individual. But standard sequencing techniques cannot always resolve which parent a group of alleles came from — known as haplotype phasing. Stanford’s application describes statistical methods for resolving the haplotype phases for a group of unrelated individuals.
The key statistical model in this technique is a hidden Markov model (HMM). In an HMM, you assume that an observable process is being influenced by something unobservable — “hidden states.” Then, by studying the process, you can learn about the hidden state. In the present case, the observable genotype is assumed to be influenced by the unobservable haplotype. By studying the genotypes of many unrelated individuals, the HMM allows estimation of the individuals’ haplotypes.
The ’982 Application
In U.S. Patent Application No. 13/486,982, Stanford describes a computerized method of estimating (inferring) haplotype phase in a collection of unrelated individuals. The ’982 Application includes two independent claims, of which Claim 1 is representative.
- A computerized method for inferring haplotype phase in a collection of unrelated individuals, comprising:
receiving genotype data describing human genotypes for a plurality of individuals and storing the genotype data on a memory of a computer system;
imputing an initial haplotype phase for each individual in the plurality of individuals based on a statistical model and storing the initial haplotype phase for each individual in the plurality of individuals on a computer system comprising a processor a memory [sic];
building a data structure describing a Hidden Markov Model, where the data structure contains:
a set of imputed haplotype phases comprising the imputed initial haplotype phases for each individual in the plurality of individuals;
a set of parameters comprising local recombination rates and mutation rates;
wherein any change to the set of imputed haplotype phases contained within the data structure automatically results in re-computation of the set of parameters comprising local recombination rates and mutation rates contained within the data structure;
repeatedly randomly modifying at least one of the imputed initial haplotype phases in the set of imputed haplotype phases to automatically re-compute a new set of parameters comprising local recombination rates and mutation rates that are stored within the data structure;
automatically replacing an imputed haplotype phase for an individual with a randomly modified haplotype phase within the data structure, when the new set of parameters indicate that the randomly modified haplotype phase is more likely than an existing imputed haplotype phase;
extracting at least one final predicted haplotype phase from the data structure as a phased haplotype for an individual; and
storing the at least one final predicted haplotype phase for the individual on a memory of a computer system.
The novelty of the ’982 Application is in the specific HMM employed. By using imputed haplotypes as the hidden states, Stanford’s method improves the accuracy of haplotype predictions. Importantly, while the claim does attempt to recite hardware limitations, the CAFC noted that “it is hard to imagine a patent claim that recites hardware limitation in more generic terms” than does Claim 1.
Stanford’s computational method of genome analysis fails the Alice test
In applying the first step of the test set forth by the Supreme Court in Alice, the CAFC agreed with the board in that the ’982 Application is directed to “mathematical calculations and statistical modeling.” This is not surprising given the nature of the claims, which even recite specific mathematical formulas. Nor did the recited steps of “receiving, … imputing, … extracting, … and storing” data convert the math into a practical application. According to the Federal Circuit, those steps are too generic and could be implemented with a regular computer.
The CAFC also agreed with the board in that the claims of the ’982 Application do not improve a technological process. In its brief and during oral argument, Stanford leaned heavily on Enfish, arguing that its claims represent an improvement to a technological process in bioinformatics. It based that claim on improvements in accuracy and efficiency of its method over the prior art. The U.S. government argued that any improvement was only to the math itself — an improvement to abstract math is still only abstract math. The court agreed that “[t]he different use of a mathematical calculation, even one that yields different or better results, does not render patent eligible subject matter.”
Finally, as noted above, the court found no inventive concept that could rescue the claims. The steps of manipulating data could be performed on a generic computer. Even considered as an ordered combination, the claimed method is only a “basic tool of scientific … work.”
First, preserve all arguments so that you can avoid waiver on appeal. At the board, Stanford failed to raise increased efficiency as a technical improvement of its method. When Stanford attempted to argue this point at the CAFC, the court refused to address it since the argument had been waived.
Second, as the U.S. Government said during oral arguments, “narrow math is still math.” No matter how specific the method, it must be more than an abstract idea. The degree of preemption by the ’982 Application would have been narrow to none. But while preemption is a mark of ineligibility, “the absence of complete preemption does not demonstrate patent eligibility.”
Finally — given the complexity and ever-changing nature of the law of subject-matter eligibility under § 101 — involve experienced counsel early. Problems with the initial application can be difficult to correct later, so it is beneficial to frame the application correctly in the first instance. Bradley’s attorneys have significant experience with subject matter eligibility and would be glad to assist with the prosecution of software- and data-related inventions.