PTAB: Protein modeling software claims are patent-eligible, arguing preemption

It has been over a year since the Federal Circuit decided McRO v. Namco Bandai. In short, McRO decided that software models can be patent-eligible. This long-awaited case sent ripples of hope to many in the patent bar, especially to those reeling from the confusion and unpredictability of Alice decided two years prior. But the McRO message has still not made its way to all examiners, especially in the life sciences. Fortunately, the PTAB sees fit to correct and overturn such examiners, as shown in a recent case Ex parte Ohrn, Appeal No. 2017-003914 (PTAB Nov. 20, 2017).

In Ex parte Ohrn, the Examiner had argued that the claims simply used a computer as a tool, i.e., “converting one form of numerical representation into another by organizing information through mathematical concepts,” and were thus an abstract idea. In addition to holding that this was an overgeneralization of the claims, the Board used three main arguments to counter this assertion.

First, the Board found that the exact manner in which the modeling was implemented was novel, and thus the Examiner failed to provide evidence that the computer was used simply as a tool for generic activity. The Board found that the “claimed method uses a combined order of specific rules that render information into a specific format that is then used and applied to create desired results.”

Second, the specific, claimed features of rules permit the improvement realized by the invention. Claim 1 focused on a specific articulated improvement in protein modeling instead of a result of an abstract idea itself.

Third, the Board found the preemption argument persuasive that the specific method recited in the claims did not improperly monopolize the basic tools of scientific and technological work. It cited fairly specific characterizations made by the appellant of what the claims do not cover.

While the preemption argument can be a strong one, the savvy patent practitioner will tell you that this argument is not without consequence. For purposes of prosecution history estoppel, characterizations made on the record as to what the claimed invention does not cover greatly weakens or even destroys such equivalence coverage in litigation. But a granted patent narrowed by such prosecution history estoppel can be better than an application where the argument was not used, leading to an abandoned application. That being said, the preemption argument can be a double-edged sword.

Another “good” bioinformatics claim found patent-eligible by the Board — insufficient evidence of routine and conventional claim

In a previous post, we defined what a good bioinformatics claim looks like. Such a claim does not suffer from divided infringement issues and it does not recite steps to be performed by hospitals or doctors (e.g., medical procedures, administration of therapies) for maximum damages. We also reported that the US Patent Office resists granting such claims, alleging a lack of statutory subject matter. Here, we report on the PTAB continuing a trend of reversing such rejections of these types of claims.

In Ex parte Shioyama, Appeal No. 2016-001637 (PTAB October 20, 2017), claims recited a cell analyzer for determining a malignancy grade of cancer. While the representative claim is an apparatus claim (i.e., cell analyzer), the claim set also included a method claim. See representative claim below.

The Board reversed the abstract idea under step 1. In its reasoning, the Board held that the Examiner failed to provide evidence to support a prima facie case of lack of patentable subject matter. Drawing support from Enfish, the Board looked to the specification to analyze the claimed invention. It agreed with the appellant that “there is something more than a mere abstract idea”.

In its holding, the Board’s overturned the Examiner’s rejection (including that the abstract idea is routine and conventional) for not being supported by sufficient objective evidence. The Board recharacterized the alleged abstract idea. Instead of “determining the malignancy of a sample by analyzing a histogram”, as asserted by the Examiner, the Board found that the claim was directed to the abstract idea of “configuring a conventional processor to obtain a number of strong area cells that are distributed in an area where the fluorescence intensity is stronger than normal cells, and determine a malignancy grade of cancer using the number of strong-area cells and the histogram.” The Board then found that in light of these elements missing from the Examiner’s asserted abstract idea, the abstract idea claim step was not routine and conventional.

The Board made it very clear that the burden to provide evidence that an alleged abstract idea is routine and conventional is squarely on the Examiner’s shoulders.

Another interesting note about this representative claim relates to the word count. The Anticipat team has been informally observing claims where the Board has reversed abstract idea rejections to see if any patterns jump out. But with the various technical fields from which these claims arise, the relatively small sample size of these claims, and diverse claim drafting styles for these claims, the most telling pattern we found for predicting a patent-eligible claim was word count. That is, the longer the claim, the more likely the Board will overturn an Examiner’s Section 101 rejection. Here, however, this claim certainly bucks this theory that a hefty word count is needed for a patent-eligible claim, the representative claim having only 85 words. See below. It turns out brevity and patent-eligibility can go hand in hand.

And while we’re on the topic of predicting patent-eligibility, yes we checked “Ask Alice” for this claim, and it outputs a 55% patent-eligible score.

Representative Claim:

1. A cell analyzer comprising:

a cytometric device which measures cells that are nuclear stained;

a display which displays a histogram of a fluorescence intensity by using a result of the measurement by the cytometric device; and

a computer comprising at least one processor configured to obtain a number of strong area cells that are distributed in an area where the fluorescence intensity is stronger than normal cells, and determine a malignancy grade of cancer using the number of strong-area cells and the histogram.

The PTAB Shows It Is Possible to Get “Good” Bioinformatics Claims Patent-Eligible

The bioinformatics market continues to show signs of impressive market growth. But patent eligibility laws in the US have not helped the cause of such a promising industry. But a recent PTAB decision, Ex parte Donnett, Appeal No. 2017-003694 (PTAB September 29, 2017), reversing the examiner’s abstract idea rejection may indicate a turning of the tide.

Before delving into the case, the subject matter eligibility issues associated with bioinformatics is not that it isn’t possible to get allowed claims. It is. The problem lies in getting good allowed claims. As a software invention, the novel feature (and often incredibly groundbreaking technological improvements) typically takes place before performing a medical procedure. For example, pioneering heart modeling software that is able to simulate where is a personalized optimal ablation target is inventive for the software piece–not for performing an ablation procedure based on the software.

Since Alice and Mayo, however, the USPTO has pushed for these “post-invention” medical procedures especially for inventions having clinical application because these medical procedure steps tie in the real world. This is despite real-world limitations not being required for other technologies. See McRO and Enfish. And in a recent CLE hosted by Fenwick & West, former SPE of art unit 1631 Marjorie Moran continued to talk about the need to tie in real-world steps in these claims.

The only problem is that reciting these additional steps creates a divided infringement nightmare for patentees, especially for bioinformatics inventions that have clinical applications. Since a doctor would be performing the medical procedure and since presumably another party could perform the software piece that precedes the medical procedure, patentees would have to pursue a difficult infringement argument. Plus, even proving infringement, Section 287 deprives a patentee the ability to collect damages against a doctor or healthcare entity.

So it appears that, while Europe, China and other countries have made strides in clarifying a reasonable patent-eligibility framework, the US has taken steps backward in the aftermath of Alice v. CLS Bank and Mayo. Bioinformatics technology are at the cross-roads of two Supreme Court decisions unfavorable to patent-eligibility: one from the software side and one from the life sciences side. And because of the relatively few patent applications for bioinformatics inventions, there is a dearth of guidance from Federal Circuit decisions directly on point. This is why the following PTAB decision ex parte Donnett is so interesting.

The claim at issue in ex parte Donnette is directed to predicting seizures. Claim 1 recites an apparatus configured to predict an upcoming neurologically abnormal state by comparing the indication of the set of correlations obtained during a sampling time period to each of Normal and Non-Normal templates.

In its analysis, the Board first found that the Examiner’s step 1 analysis was erroneous. Just because the claim recites correlations, this does not mean that the Examiner had sufficiently shown that the claim is directed to an abstract idea ‘of itself’. The Board instead quoted the appellant’s argument that the claims are directed to seizure prediction, with an innovation in digital EEG signal processing. The Board quoted the appellant’s argument that “the innovation reflects both an improvement in the functioning of the signal processor and an improvement in another technology, seizure prediction.”

At the conclusion of its decision, the Board seemed a bit non-committal. That is, the Board did not expressly conclude that the claimed invention was not directed to an abstract idea because of the technological improvements articulated by the appellant. It merely quoted the appellant’s arguments. Instead, the Board concluded that the Examiner did not sufficiently show that the conclusory statement met the minimum threshold for establishing that the claims are directed to an abstract idea. Even with such non-commitment, it is a big win for the appellant. These claims do not appear to suffer from the divided infringement issues noted above and will likely receive a Notice of Allowance shortly.

Is this the first of a tide of bioinformatics claims that are both good and patent-eligible? Perhaps. But too early to tell.