Pacific Biosciences of California, Inc. v. Personal Genomics Taiwan, Inc.

89 F.4th 1377

• Court: Court of Appeals for the Federal Circuit
• Decided: January 9, 2024
• Docket: 22-1410
• Status: Published

Opinion

Case: 22-1410 Document: 52 Page: 1 Filed: 01/09/2024

United States Court of Appeals for the Federal Circuit ______________________

PACIFIC BIOSCIENCES OF CALIFORNIA, INC., Appellant

v.

PERSONAL GENOMICS TAIWAN, INC., Cross-Appellant ______________________

2022-1410, 2022-1554 ______________________

Appeals from the United States Patent and Trademark Office, Patent Trial and Appeal Board in Nos. IPR2020- 01163, IPR2020-01200. ______________________

Decided: January 9, 2024 ______________________

EDWARD R. REINES, Weil, Gotshal & Manges LLP, Red- wood Shores, CA, argued for appellant. Also represented by DEREK C. WALTER.

KEITH ORSO, Irell & Manella LLP, Los Angeles, CA, ar- gued for cross-appellant. Also represented by ALAN J. HEINRICH; MICHAEL RICHARD FLEMING, Washington, DC. ______________________

Before PROST, TARANTO, and HUGHES, Circuit Judges. TARANTO, Circuit Judge. Case: 22-1410 Document: 52 Page: 2 Filed: 01/09/2024

Pacific Biosciences of California, Inc. (PacBio) filed two petitions with the Patent and Trademark Office under 35 U.S.C. §§ 311–19, each one seeking an inter partes review of a group of claims of U.S. Patent No. 7,767,441, which is owned by Personal Genomics Taiwan, Inc. (PGI). The Pa- tent Trial and Appeal Board, acting for the PTO’s Director, instituted both IPRs, which overlapped in the claims chal- lenged but differed in the prior art invoked. The Board eventually issued final written decisions in the IPRs. In one of the IPRs, the Board rejected PacBio’s challenge to claims 1–2, 6–7, 10–22, 24, and 27–36. Pacific Biosciences of California, Inc. v. Personal Genomics Taiwan, Inc., No. IPR2020-01200, 2022 WL 214042 (P.T.A.B. Jan. 18, 2022) (’1200 Decision). In the other IPR, the Board agreed with PacBio’s challenge to claims 1–6, 9, and 43–58. Pacific Bi- osciences of California, Inc. v. Personal Genomics Taiwan, Inc., No. IPR2020-01163, 2022 WL 212276 (P.T.A.B. Jan. 18, 2022) (’1163 Decision). Under those two decisions, claims 7, 10–22, 24, and 27–36 survive; claims 1–6, 9, and 43–58 do not. Both parties appeal. Appellant PacBio principally challenges the Board’s construction of the claim phrase “identifying a single biomolecule,” while also briefly chal- lenging the Board’s finding that the prior art PacBio in- voked in the ’1200 IPR to meet this limitation does not teach it under the Board’s construction. Cross-appellant PGI, besides defending the Board’s construction of the dis- puted claim phrase, challenges the Board’s factual findings that the PacBio-invoked prior art in the ’1163 IPR teaches the disputed claim phrase. We affirm both decisions. I U.S. Patent No. 7,767,441 describes and claims an “ap- paratus for identifying a single biomolecule” as well as methods of using or making that apparatus. ’441 patent, col. 26, line 11; see also id., col. 26, line 10 through col. 30, line 5. The patent describes an apparatus that uses many Case: 22-1410 Document: 52 Page: 3 Filed: 01/09/2024

PACIFIC BIOSCIENCES OF CALIFORNIA, INC. v. 3 PERSONAL GENOMICS TAIWAN, INC.

“optical detection apparatuses” to “monitor a large number (e.g., in some embodiments, more than 10,000) of single bi- omolecules in parallel,” and thereby determines the iden- tity of many biomolecules in a sample “with high throughput.” Id., col. 3, lines 55–65; see also id., col. 4, lines 11–16. The “optical detection apparatus” uses a “light de- tector” that is in close proximity to (e.g., “less than or equal to 100 micrometers” from) a “linker site” that is “treated to affix the biomolecule” to be identified. Id., col. 2, lines 30– 44. The light detector can measure a signal from some light-emitting molecule—e.g., “a fluorophore attached to the biomolecule,” a “labeled probe,” or “labeled nucleo- tides”—and thereby identify the affixed biomolecule. Id., col. 2, lines 43–62; see also id., col. 17, lines 58–62 (discuss- ing “chromophores”); id., col. 18, lines 18–21 (same). Claim 1 is representative for present purposes: 1. An apparatus for identifying a single biomole- cule, comprising: a substrate having a light detector; and a linker site formed over the light detector, the linker site being treated to affix the bi- omolecule to the linker site; wherein the linker site is proximate to the light detector and is spaced apart from the light detector by a distance of less than or equal to 100 micrometers. Id., col. 26, lines 11–18. In the ’1200 IPR, PacBio challenged claims 1–2, 6–7, 10–22, 24, and 27–36—all claiming the apparatus of claim 1 or its use and many, though not all, focusing specifically on nucleic acids and determining their nucleotide se- quences—as unpatentable for anticipation or obviousness based principally on the Hassibi reference, a published United States patent application, U.S. Patent Application Case: 22-1410 Document: 52 Page: 4 Filed: 01/09/2024

Publication No. 2004/0197793 A1 (filed Jul. 24, 2003) (pub- lished Oct. 7, 2004) (Hassibi); J.A. 1638–1706. The Board issued a final written decision determining that PacBio had not shown any of the challenged claims to be unpatent- able. ’1200 Decision, at *21. In the ’1163 IPR, PacBio challenged claims 1–6, 9, and 43–58—which refer to biomolecules generally, not to nu- cleic acids specifically—as unpatentable for anticipation or obviousness based principally on the Choumane reference, an international patent application, PCT Application Pub- lication No. WO 2007/045755 A1 (filed Oct. 17, 2006) (pub- lished Apr. 26, 2007) (Choumane); J.A. 5533–62. The Board issued a final written decision determining that Pac- Bio had proved all the challenged claims to be unpatenta- ble. ’1163 Decision, at *26. In reaching its decisions, the Board adopted a claim construction of the preamble phrase “identifying a single biomolecule,” setting forth reasoning that is materially identical in the two opinions. Compare ’1200 Decision, at *6–11, with ’1163 Decision, at *6–11. In a conclusion not challenged on appeal, the Board determined that the pre- amble phrase is a limitation on the claimed subject matter because it provides antecedent basis for references to “the biomolecule” in the body of the relevant claims. See ’1200 Decision, at *7, *9 (claim 1); ’1163 Decision, at *7, *9 (same). 1 In another conclusion not challenged on appeal, the Board determined that the full phrase “for identifying a single biomolecule” refers to a capability of the appa- ratus. See ’1200 Decision, at *9–10; ’1163 Decision, at *9–

1 All independent claims either require the appa- ratus of claim 1 (claims 11, 12, 16, 30, 43), whether ex- pressly or indirectly, or have the same relevant preamble language tied to “the biomolecule” language in the body (claims 48, 53–55). See ’441 patent, col. 26, line 10 through col. 30, line 5. Case: 22-1410 Document: 52 Page: 5 Filed: 01/09/2024

PACIFIC BIOSCIENCES OF CALIFORNIA, INC. v. 5 PERSONAL GENOMICS TAIWAN, INC.

10 (same); see ParkerVision, Inc. v. Qualcomm Inc., 903 F.3d 1354, 1361–62 (Fed. Cir. 2018) (explaining that capa- bility is one meaning of “for” language). The claim-construction dispute central to the appeals before us involves the Board’s understanding that the “identifying a single biomolecule” phrase, in the context of the “[s]pecification of the ’441 patent,” “contemplates run- ning myriad optical detection apparatuses in parallel to de- tect a single or individual biomolecule in each such apparatus.” ’1200 Decision, at *8; ’1163 Decision, at *8. That construction requires that the apparatus have the ca- pability to characterize (determine the identity of) a bio- molecule by examining that biomolecule alone, with no copies created to form an ensemble for examination. In adopting that construction, the Board rejected PacBio’s ar- gument that an apparatus would come within this claim phrase if the apparatus, though not capable of characteriz- ing a biomolecule by examining it alone, had the capability to characterize a biomolecule by making copies, examining the resulting ensemble, and inferring the identity of the starter biomolecule. ’1200 Decision, at *8 (rejecting Pac- Bio’s construction as “inapposite when the challenged claims are read in light of the [s]pecification of the ’441 pa- tent”); ’1163 Decision, at *8 (same). In the ’1200 Decision, the Board found that PacBio failed to establish that this limitation was taught by the Hassibi reference—on which PacBio relied for this essen- tial point in the only ground raised in PacBio’s appeal to us. ’1200 Decision, at *12–17. That determination sufficed to reject PacBio’s challenge to the claims at issue in that IPR. Id. at *21. In the ’1163 Decision, the Board found, in contrast, that PacBio did establish that the “identifying a single biomole- cule” limitation was taught by the Choumane reference, i.e., that Choumane taught an apparatus with the capabil- ity required by the Board’s claim construction. ’1163 Case: 22-1410 Document: 52 Page: 6 Filed: 01/09/2024

Decision, at *12–16. The Board added, id. at *22, that this limitation was also taught by a second reference that Pac- Bio invoked for obviousness in combination with Choumane, namely, a published patent application, U.S. Patent Application Publication No. 2002/0182716 A1 (filed Feb. 12, 2002) (published Dec. 5, 2002) (Weisbuch); J.A. 5563–80. The Board found the requisite proof as to the other limitations of the challenged claims, by Choumane alone or in the specified combinations for obviousness, but those findings are not at issue on appeal. The Board there- fore found that PacBio had shown unpatentability of the claims challenged in the ’1163 IPR. ’1163 Decision, at *26. The Board issued both of its final written decisions on January 18, 2022. PacBio filed a timely notice of appeal from the ’1200 Decision on January 24, 2022, and PGI filed a timely notice of appeal from the ’1163 Decision on March 18, 2022. The appeals are authorized by 35 U.S.C. §§ 141(c) and 319, and we have statutory jurisdiction under 28 U.S.C. § 1295(a)(4)(A). PGI’s pending infringement suit against PacBio under the ’441 patent supports constitu- tional standing. See PacBio’s Opening Brief at 1. II A PacBio’s appeal principally challenges the Board’s con- struction of the “identifying a single biomolecule” claim limitation. We review this construction, which rests on in- trinsic evidence, without deference. Polaris Innovations Ltd. v. Brent, 48 F.4th 1365, 1372 (Fed. Cir. 2022). We af- firm the Board’s construction of “identifying a single bio- molecule” as requiring an apparatus capable of ascertaining the identity of one single, individual biomole- cule by examining only that biomolecule. It is undisputed before us that the claim language at issue—“for identifying a single biomolecule”—refers to a capability of an apparatus. What is disputed is the Case: 22-1410 Document: 52 Page: 7 Filed: 01/09/2024

PACIFIC BIOSCIENCES OF CALIFORNIA, INC. v. 7 PERSONAL GENOMICS TAIWAN, INC.

meaning of “identifying a single biomolecule.” That phrase has an ordinary meaning on its face and in context. The language refers to (a) ascertaining the identity of a biomol- ecule, i.e., what that biomolecule is, and (b) doing so by ex- amining just that one biomolecule, not others (even copies). It is that capability the apparatus must have, no matter what other capabilities it has or how the apparatus may be used in a particular instance. The proper construction in this case is determined by the ordinary meaning of the claim language understood in the context of the patent document. See World Class Tech- nology Corp. v. Ormco Corp., 769 F.3d 1120, 1123 (Fed. Cir. 2014); Phillips v. AWH Corp., 415 F.3d 1303, 1312–17 (Fed. Cir. 2005) (en banc). The ascertainment of the identity of a molecule is the ordinary meaning in this context of ‘iden- tify,’ as the aim of the invention is to ascertain the identity, e.g., the nucleotide sequence, of the single biomolecule. See BRYAN A. GARNER, GARNER’S MODERN AMERICAN USAGE 435 (3d ed. 2009) (defining “identify” as “(2) to ascertain or demonstrate what something . . . is”); RANDOM HOUSE WEBSTER’S UNABRIDGED DICTIONARY 950 (2d ed. 2001) (de- fining “identify” as “1. to recognize or establish as being a particular . . . thing; verify the identity of”). That aspect of the claim meaning is not in genuine dispute. Only the above-stated second part of the meaning is disputed. We agree with the Board that this identifying-by-ex- amining-one-alone meaning is the ordinary meaning of the phrase in context. The striking feature of the phrase is its inclusion of the word “single.” There is no apparent reason for the inclusion of the word “single” in the phrase except to indicate that the capability required is to identify a mol- ecule with just that one molecule in view. The ’441 patent’s specification confirms this under- standing by repeatedly stressing that this “single biomole- cule” capability is critical to the invention. It states that it is describing “a bioassay system for identifying a single Case: 22-1410 Document: 52 Page: 8 Filed: 01/09/2024

biomolecule at a detecting unit.” ’441 patent, col. 2, lines 30–31. In such a bioassay system, the specification indi- cates, “[e]ach optical detection apparatus may sense the ex- istence of a fluorophore on the single molecule by detecting photons emitted from the fluorophore.” Id., col. 3, lines 59– 61. When the patent illustrates optical detection apparat- uses “identifying a single biomolecule,” it depicts examina- tion of one individual biomolecule, not an ensemble or cluster consisting of multiple biomolecules. See id., figs. 2 & 6–8. Significantly, in explaining the claimed bioassay sys- tem’s single-molecule sensitivity, the specification differen- tiates, on one hand, “identifying a single biomolecule” by examining that individual biomolecule from, on the other, detecting a population-level signal from an ensemble or cluster of amplified or copied biomolecules. It indicates that the latter process involves problems that the former avoids. In particular, the patent requires that the claimed apparatus be capable of the former functionality, explain- ing that the claimed “devices should be capable of sequenc- ing single molecules to avoid the known difficulty of asynchrony in both the amplification (e.g., drift between the sequences of ideally clonal templates) and sequencing (e.g., dephasing of the stepwise sequencing reactions amongst the sequencing templates) steps of clustered se- quencing methods.” Id., col. 1, line 67 through col. 2, line 6. Avoiding those problems was accomplished by an appa- ratus with single-molecule detection sensitivity, enabling single-molecule examination for identification, confirming the proper understanding of the claim language here in dis- pute. Other claims of the patent—method claims that call for particular uses of the claim 1 apparatus—provide support for this claim construction. Method claim 26 depends on claim 16, which claims a method (using a claim 1 appa- ratus) of “performing nucleic acid sequencing” of “one nu- cleic acid molecule,” but adds the limitation “wherein the Case: 22-1410 Document: 52 Page: 9 Filed: 01/09/2024

PACIFIC BIOSCIENCES OF CALIFORNIA, INC. v. 9 PERSONAL GENOMICS TAIWAN, INC.

nucleic acid is amplified.” ’441 patent, col. 27, lines 20–21 (claim 26); id., col. 26, lines 59–64 (claim 16). Claim 30 claims a method (using a claim 1 apparatus) of “detecting” “one or more biomolecule.” Id., col. 27, lines 30–34. Those claims thus call for using the claim 1 apparatus for ex- pressly described multiple-molecule examinations, includ- ing where amplification of a starter molecule produces the group examined, implicitly requiring a capability to do so. Dependent claims often add capabilities to those stated in the independent claim (where there is no inconsistency), and here the language used underscores what is not re- quired by the claim 1 phrase itself. We reject PacBio’s contention that the Board’s (and our) reading “conflates identifying a single molecule with detecting a single molecule.” PacBio’s Reply and Re- sponse Brief at 2. We need not deny that, in some contexts, one might say that a molecule can be identified by following processes for making copies, examining the resulting en- semble, and inferring the identity of the starting molecule. But the ’441 patent recognizes the problems with such pro- cesses and seeks to avoid those problems by claiming an apparatus capable of identification by single-molecule ex- amination. Such a claim does not “conflat[e]” identification with detection but instead uses a detection requirement to specify a particular identification capability. For those reasons, we affirm the Board’s construction of the claim language in dispute, understood as we have explained it. B Both PacBio and PGI challenge the Board’s factual findings regarding the prior art. PacBio challenges the Board’s finding in the ’1200 IPR that Hassibi does not dis- close “identifying a single biomolecule” under the claim construction we are affirming. PGI challenges the Board’s finding in the ’1163 IPR that Choumane does disclose “identifying a single biomolecule” under that construction. Case: 22-1410 Document: 52 Page: 10 Filed: 01/09/2024

We conclude that both findings are supported by substan- tial evidence, and we therefore affirm them. See Shoes by Firebug LLC v. Stride Rite Children’s Group, LLC, 962 F.3d 1362, 1369 (Fed. Cir. 2020) (explaining that we review Board findings regarding the content of prior art for sub- stantial-evidence support). 2 1 In the ’1200 IPR, the critical disclosure in Hassibi is that the principal embodiment of the described and claimed invention uses an assay—a bioluminescence re- generative cycle (BRC) assay—that Hassibi describes as having “a sensitivity of detection as low as 0.1 attomoles.” J.A. 1670–71 ¶¶ 16–17; see also J.A. 1704, Claim 24 (“wherein sensitivity of detection is at least 0.1 attomol”). This is a minimum-detection limit of greater than 60,000 molecules. J.A. 3800, lines 1–4 (testimony of PacBio ex- pert). This assay is referred to repeatedly in Hassibi for detection purposes, and in other places, Hassibi discloses an even higher minimum sensitivity. J.A. 1696 ¶ 325 (re- porting a sensitivity of “1 amol to 100 amol”). The Board credited the testimony of Dr. Harris (PGI’s expert) and Hassibi itself in finding that nothing in Hassibi or else- where in the record rebuts this evidence by demonstrating that Hassibi, instead, discloses the capacity to ascertain the identity of one individual biomolecule using either a more sensitive BRC assay or a different assay. ’1200 Deci- sion, at *15–17. This evidence provides substantial-evi- dence support for the Board’s finding that Hassibi does not disclose “identifying a single biomolecule.”

2 Because we affirm the Board’s finding about the teaching of Choumane, we need not address PGI’s chal- lenge to the Board’s finding about the teaching of Weis- buch. Case: 22-1410 Document: 52 Page: 11 Filed: 01/09/2024

PACIFIC BIOSCIENCES OF CALIFORNIA, INC. v. 11 PERSONAL GENOMICS TAIWAN, INC.

2 The critical disclosure in Choumane is that the claimed invention can include “very small openings . . . of a dimen- sion less than the wavelength of light emitted by chromo- phores,” where “[t]hese openings delimit very small observation volumes . . . for the detection and observation of individual chromophores.” J.A. 5551, lines 20–24 (em- phasis added). The Board reasonably understood this to disclose the capacity to identify one individual biomolecule, given that a single chromophore is often used to tag a single biomolecule. ’1163 Decision, at *15–16; J.A. 5360 ¶ 61; ’441 patent, col. 3, lines 59–61. Added support is found in Choumane’s description of “a biosensor with ultrasensitive detection” for detecting “biological probes,” not merely chromophores in isolation. J.A. 5538, lines 3–6. PGI’s principal challenge to the Board’s finding relies on a calculation, starting from (and perhaps extending a bit beyond) analysis supplied by its expert witness, that Choumane’s disclosures do not disclose a detection sensi- tivity of less than 78 biomolecules. PGI’s Principal and Re- sponse Brief at 68–73 (citing J.A. 7605–06 ¶ 86). Although PacBio seems to suggest otherwise, this type of calculation can be proper where based on the express disclosures of a piece of prior art. See Perfect Web Technologies, Inc. v. In- foUSA, Inc., 587 F.3d 1324, 1329 (Fed. Cir. 2009) (“We therefore hold that while an analysis of obviousness always depends on evidence that supports the required Graham factual findings, it also may include recourse to logic, judg- ment, and common sense available to the person of ordi- nary skill that do not necessarily require explication in any reference or expert opinion.”). In this case, however, the Board, considering the calculation, had sufficient reason, in light of the Choumane statement quoted above, to reject the inference about Choumane that PGI offered the calcu- lation to support. Case: 22-1410 Document: 52 Page: 12 Filed: 01/09/2024

PacBio’s expert provided rebuttal testimony stating that Choumane teaches “improv[ing] collection efficiency by a factor of 400,” which, PacBio’s expert ultimately con- cluded, allows a sensitivity increase over the prior art suf- ficient to disclose single-biomolecule detection capability. J.A. 7427–28 ¶ 6; see also J.A. 7428 ¶ 7 (calculating that Choumane discloses “a minimum detection of 0.975 chro- mophores (i.e. single molecule detection)”); cf. J.A. 7428–29 ¶ 8 (calculating that combining Choumane’s disclosures with Weisbuch’s results in “a minimum detection of 0.25 chromophores”). In contrast, PGI’s expert stated that when Choumane expressly discloses improving sensitivity over the prior art only by “several tens of times,” J.A. 5537, line 3, it means only about “30 or 40 or 50” times better, J.A. 7334, line 2, which, PGI has argued, is not sufficient for single-biomolecule detection sensitivity. See PGI’s Re- ply Brief at 25 n.4 (“Substituting a 40-fold improvement for [PacBio’s expert’s] 400-fold improvement would increase sensitivity . . . [to] about 9.5 chromophores.”). Where the overall evidence reasonably allows the Board’s factual find- ing on a point, we do not “reweigh the evidence” to reject that finding. Regents of the University of California v. Broad Institute, Inc., 903 F.3d 1286, 1294 (Fed. Cir. 2018). Here, on the evidence before it, the Board could reasonably credit PacBio’s expert, ’1163 Decision, at *16, and rely on the “detection and observation of individual chromophores” language of Choumane, J.A. 5551, line 24, for its finding about Choumane’s teaching. III We find none of PacBio’s and PGI’s remaining argu- ments persuasive. The Board’s final written decisions in the ’1200 IPR and the ’1163 IPR are affirmed. The parties shall bear their own costs. AFFIRMED