About the Authors

  • The Authors and Contributors of "Patent Docs" are patent attorneys and agents, many of whom hold doctorates in a diverse array of disciplines.
2018 Juristant Badge - MBHB_165
Juristat #4 Overall Rank

E-mail Newsletter

  • Enter your e-mail address below to receive the "Patent Docs" e-mail newsletter.

Contact the Docs


  • "Patent Docs" does not contain any legal advice whatsoever. This weblog is for informational purposes only, and its publication does not create an attorney-client relationship. In addition, nothing on "Patent Docs" constitutes a solicitation for business. This weblog is intended primarily for other attorneys. Moreover, "Patent Docs" is the personal weblog of the Authors; it is not edited by the Authors' employers or clients and, as such, no part of this weblog may be so attributed. All posts on "Patent Docs" should be double-checked for their accuracy and current applicability.
Juristat #8 Overall Rank


« Patent Luminaries Try to Set Congress Straight on Drug Price Controls | Main | Overcoming the Consequences of In re Cellect »

October 02, 2023


Thank you for this post. I think the alternatives discussed are the most likely to work, but am mindful that a number of examiners in Art Unit 1600 have rejected any claim language directed to allowing some even small amount of sequence variation (including for example sequences with very high (99%) percent identity or sequences varying by "less than" a small number of amino acids or nucleotides in a molecule with known conserved regions). Examiners have also rejected claims encompassing the *use* of antibodies that bind to a known antigen even when there are several well established examples in the art, which seems at odds with Capon v Eshhar (although maybe that case has now been made obsolete). So unfortunately I think in practice it may now be nearly impossible to get claims of any useful breadth.

I think Kevin's post is, as usual, brilliant. Also, recently, Sharon Crane of Rothwell Figg gave a brilliant treatment of this written description topic, and Joanna Brougher of the BioPharma Law group, masterfully dealt with MPF, both on AIPLA's September 20 Hot Topics in Chem and Pharma program. For those who missed it, recordings are available through AIPLA. I stand strongly for MPF, having spent years promulgating it. To be sure, what exactly is a literal equivalent will depend on the specific facts and circumstances. In the pharma space, I am aware of only one case, Wockenhardt, where a pharma MPF claim was litigated. The case settled, but what caught my eye was that the defendant accepted an injunction as part of the settlement. Before I retire in 2066, I expect we will have much more light and knowledge about MPF in pharma. I am using it with my pharma clients at The Marbury Law Group and my former law firm has done the same.

Dear Darden:

While I appreciate your bleak outlook, I think the only way to address the issue is to abandon the "picnic blanket" approach to claim scope and adopt a "picket fence" approach (or perhaps a pointillist approach) in identifying as many related sequences that can be verified so that the "millions and millions" meme is not applicable. (I'm sure the sequence comparison software will identify the number of permutations in any consensus sequence.) Coupled with the routine nature of the binding tests to verify function based on structure at the minimum competitors will need to invest the time and effort to design around (which is sufficiently significant that the "first mover" advantage may be enough to make the effort worthwhile).

Thanks for the comment.

Dr. Noonan, this is outside of my practice area, but a potential overlap with an area that I am familiar with occurs with your comment of, "the sequence comparison software will identify the number of permutations in any consensus sequence."

If - and this has been admitted to by at least one pharma company in public discussions on Artificial Intelligence (AI) - the "identifying" action is by 'software,' AND that is the action that is setting out the patentable distinction, then you will have a circumstance in which an invention is being made by a non-human actor (the AI, in at least a co-inventor role), and no viable patent will be forthcoming -- either by the prong of no patents for non-human inventors OR the prong that anything from an AI may well be considered obvious [if AI's are deemed non-inventive].

Further, you advance a notion of "first mover" advantage, but I think such a notion is still-born given the (same) burgeoning use of AI for speed of development.

Again, pharma is typically outside of my bailiwick, but it does not take operating in that sphere to see some serious oncoming issues for an industry that has so heavily depended on the patent model to eek out payback (and often in the later days of patent coverage).

Dear Skeptical:

Going in inverse order, the "first mover" idea is that the hypothetical patentee will have actually made several antibodies, subjected them to testing and development, and obtained FDA approval to market them. Given the effort all that requires (which will also be required by copycats and design arounds) there is likely to be a significant lag in competing antibodies being marketed, and that is the "first mover" effect I posit might give an advantage.

As for AI, in this instance it is just a tool, because it is not creating anything but the comparison, which is based on the sequence of the antibodies themselves that are real world molecules. Producing the comparison does not rely on any of the properties that make AI intelligent; indeed, such comparisons have been routine for ~ 40 years.

Thanks for the thoughts.

Thank you Dr. Noonan - I think that you are not addressing the novel characteristics of the use of AI, which brings it to be MORE than merely using a tool.

IF in fact, one wants to depend on "just using a tool,"
THEN this necessitates the view that the tool is NOT providing any 'inventive step' and the results of the use of the tool is de facto obvious -- obvious to a NEW, albeit still non-real person but instead legal fiction) Person Having Ordinary Skill In The Art.

There is a Hobson's choice in the current "mere use of" AI that needs be addressed.

I too have long been an advocate of MPF claims in biology (though not nearly as prominently as Tom) and completely agree with him that exactly what an equivalent will be in the view of the courts is an open question. However, it has to be broader than simply claiming the disclosed sequences. What value that has will likely vary widely from case to case. Kevin worries about the testing that will be necessary to confirm an equivalent is in fact one and suggests pursuing consensus sequences instead. By all means, construct a consensus sequence. No one claim type will be a panacea. But please realize that there will still be testing required to determine if any particular sequence encompassed by the consensus sequence has the associated function expected.

Dear Skeptical: I have no trouble confessing ignorance about AI. I just don’t think it’s relevant because (to make an analogy) what I suggest can be performed with an abacus and doesn’t need a digital computer.

And I am discussing sufficiency of disclosure - the novelty and nonobviousness derive from the biology and its application (I recognize the risk of the section 101 folks making an eligibility argument).


The comments to this entry are closed.

June 2024

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29