The transcendental conundrum in patent law in these times is how to overcome the misinterpretation of the Supreme Court's decisions on patent eligibility law by district courts and the Federal Circuit.  That these courts cannot overcome the precedential tangle they have created is firmly established by the Court's own decisions; Judge Moore in Athena Diagnostics, Inc. v. Mayo Collaborative Services, LLC stated the obvious when she said in her dissent:

My colleagues' refusal deflates the Amici's hopeful suggestion that our precedent leaves the eligibility of a diagnostic claim in front of the Federal Circuit "uncertain."  It is no longer uncertain.  Since Mayo, every diagnostic claim to come before this court has been held ineligible.  While we believe that such claims should be eligible for patent protection, the majority of this court has definitively concluded that the Supreme Court prevents us from so holding.  No need to waste resources with additional en banc requests.

And in response, the Supreme Court has eschewed any and every opportunity to revisit their twisted precedent denying certiorari in over 50 cases (see "U.S. Supreme Court on Eligibility: Nothing to See Here, Move Along").

Part of the problem is that it has been maddeningly difficult to define not what patent eligibility is (you cannot go wrong with "anything under the sun made by man") but rather what it is not.  In the high technology class of inventions, this has come down to deciding without defining what an abstract idea is and when its abstractness prevents patent eligibility; see, e.g., "Stupid §101 Tricks").  (The other, related type of ineligible subject matter are business method patents, the exclusion of which is almost categorical (see "Bilski v. Kappos, Alice Corp. Pty. Ltd. v. CLS Bank Int'l"; "CyberSource Corp. v. Retail Decisions, Inc."); this has the benefit is requiring little interpretation and hence maximal certainty regarding what is ineligible.)

But the ineligibility of the latest iGadget, while sometimes tragic, is not as existentially problematic as the havoc that these precedents have wreaked on life sciences patenting.  For both diagnostic methods and to a slightly lesser extent natural products, the philosophically lost proscriptions by the Court, bolstered by plain illogic in district court (see "Ariosa Diagnostics, Inc. v. Sequenom, Inc.") and Federal Circuit (see "Federal Circuit Denies Rehearing en banc in Ariosa v. Sequenom") decisions, has rendered pursuit of patent protection for these inventions to be relegated to the ranks of the foolhardy.  The effect on investment and hence progress and innovation has been as expected; perhaps the only silver lining from the SARS-Cov-2 epidemic has been that in the frantic and desperate struggle for both diagnostics and vaccines the usual market forces have been collapsed by government investment (which is not usually a recipe for economic success).

But if policymakers cannot find a path to resolve the political differences that have prevented a legislative solution to the patent eligibility issue (see "The Zombie Apocalypse of Patent Eligibility Reform and a Possible Escape Route"), perhaps it is because the various stakeholders have not been able to arrive at a consensus over what should and should not be eligible (besides interest groups like the ACLU shouting "hands off my DNA"; see "ACLU (Predictably) Opposes Patent Subject Matter Eligibility Proposal").  This situation suggests it might be better to start with examples of what should be patent eligible (or at least something that would focus the basis for opposition to eligibility) and walk backwards to the point where various stakeholder groups could raise objections that might define the metes and bounds of why certain inventions have or should be deemed unprotectable under patent law.

For diagnostic methods, it might be best to start with that most tangible of subject matter classes, a machine (see "American Axle v. Neapco Holdings" and "Ex parte Itagaki and Nishihara" perhaps to the contrary).  In this example, the claimed invention would be a machine configured to perform a diagnostic assay on a biological sample, wherein the machine provides a diagnosis based on detecting the presence of a naturally occurring biomarker in the sample.  The specifications on the machine can be limited so that each embodiment of the machine is calibrated to perform only specific diagnostic assays on specific samples to detect specific biomarkers that are specific for only a specific disease or class of diseases.  As so limited, should this machine be patent-eligible?  Why or why not?  And in making these determinations of course the patent eligibility of other diagnostic machines, from the MRI to the EKG to the stethoscope should guide the eligibility decision.

If this machine passes the eligibility test, the next-level question could be whether the processes used by the machine (e.g., the chemistry) is patent-eligible, and does eligibility depend on whether these assays are novel?  What is the significance of whether they are adapted to detecting the biological phenomenon informative for the diagnosis?  And what if they are specifically adapted (along with all the specificities considered above regarding patent eligibility for the machine) to use with the machine?

Another consideration is whether these processes need to be performed by the machine or whether they could be performed outside the machine itself?  Would eligibility of the diagnostic assay depend on claim infringement requiring use of the machine, i.e., producing the "inputs" from the biological sample cannot be interpreted unless evaluated by the machine?

And does the complexity of the biomarker inputs make a difference, i.e., should it matter if the machine evaluates the input information using a computer, for example?  Does it matter if the inputs are so complex they can only be understood using a computer?

The answers to these questions can then be used to evaluate the distinction(s) between these features/characteristics of these different embodiments of the invention that vitiate patent-eligibility.  In other words, where do we draw the line?

For natural products, the issue may be less complex, based on the distinction drawn by the Court in Myriad that "mere" isolation of genomic DNA rendered that DNA ineligible for patenting, and that conversion of naturally occurring messenger RNA (mRNA) to complementary DNA (cDNA) involved sufficient human intervention to render cDNA patent-eligible.  But as a general rule, this distinction with other natural products is not satisfactory for at least two reasons.  First, it flies in the face of established practice that "mere" isolation could pass the patent eligibility test.  And second, because natural products in their natural state are not usually useful, their utility arising from isolation (and purification, and concentration) and use in applications for which they are not used in nature (the canonical example being the anticancer drug taxol which, while intrinsically having anticancer properties does not exhibit this capability naturally because trees don't get cancer).

The issue may be as simple as recognizing (as District Court Judge Sweet did in his opinion invalidating Myriad's claims) that DNA is sui generis, the "physical embodiment of genetic information."  "Mere" isolation of DNA was (perhaps) less relevant because the function of the DNA was unchanged by its isolation; before this statement enrages biotechnologists it should be recalled that all the ways that isolated DNA has been used remain patent-eligible (incorporated into vectors, introduced into heterologous cells, producing valuable biological molecules, etc.).  For almost every other natural product, however, its utility is not found in nature, so perhaps that transformation, produced by human intellect, should be enough to render most important "natural" products, converted most unnaturally, to be patent eligible.  After all, a claim to "Penicillin" per se has never been patent-eligible because it reads on penicillin as it exists in nature, and a claim to "a pharmaceutical composition comprising penicillin [at a useful dose and formulated for pharmaceutical use]" has always been (and still is) patent eligible.  What is uncertain today is whether a claim to "an isolated/purified preparation of penicillin [having a specific utility, claimed or disclosed in the specification]" is patent eligible, and the outcome of this thought experiment on this subject matter is to determine why or why not?

It may only be by "walking backwards" from exemplars of these different technologies that must be patent-eligible to the point where that eligibility can be called into question that some insight may be gained as to why some subject matter (or embodiments thereof) are patent-eligible sheep and others patent-ineligible goats.  Almost a decade of letting courts try to come to these distinctions in their case law has failed to arrive at any clarity, Blackstone be damned; maybe it's enough to make the case for a more analytical approach to recognize that the result can't be any worse.