PAN et al.v.Flusberg et al.

Patent Trials and Appeals Board

Sep 3, 2014

12613291 - (J) (P.T.A.B. Sep. 3, 2014)

BoxInterferences@uspto.gov
Telephone: 571-272-4683 Entered: August 3, 2014

UNITED STATES PATENT AND TRADEMARK OFFICE
________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
________________

Patent Interference No. 105,970 (JGN)

INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE
and Ti-Shiue Biotech, Inc.
Junior Party
(US 8,846,630)

v.

PACIFIC BIOSCIENCES
of California, Inc.
Senior Party
(13/633,673 and 13/930,178)
________________

JUDGMENT
Bd.R. 127

________________

Before SALLY GARDNER LANE, JAMES T. MOORE, and
JOHN G. NEW, Administrative Patent Judges.

NEW, Administrative Patent Judge
The Board determined that all of the claims of Junior Party Industrial
Technology Research Institute’s (“ITRI”) US Patent No. 8,486,630 B2 are
unpatentable under 35 U.S.C. § 103(a).
1
Accordingly there is no basis to
continue the interference. It is—

ORDERED that judgment be entered against Junior Party ITRI for
count 1
2
;

FURTHER ORDERED that claims 1-28 of ITRI’s involved U.S.
Patent No. 8,486,630 B2 be CANCELED, 35 U.S.C. 135(a)
3
; and

FURTHER ORDERED that a copy of this judgment be entered in the
administrative records of the involved 8,486,630 B2 patent and application.

NOTICE: “Any agreement or understanding between parties to an
interference, including any collateral agreements referred to therein, made in
connection with or in contemplation of the termination of the interference,
shall be in writing and a true copy thereof filed in the Patent and Trademark
Office before the termination of the interference as between the said parties
to the agreement or understanding.” 35 U.S.C. 135(c); see also Bd.R. 205
(settlement agreements).

1
Paper No. 167
2
Paper No. 1
3
As was in effect on March 15, 2013. See Pub. L. 112-29, § 3(n),
125 Stat. 284, 293 (2011).
Attorney for ITRI:

David S. Forman, Ph.D.
M. Paul Barker
Adam M. Breier, Ph.D.
Finnegan, Henderson, Farabow, Garrett & Dunner, LLP
David.forman@finnegan.com
Paul.barker@finnegan.com
Adam.breier@finnegan.com

Attorney for PacBio:

Michael S. Tuscan, Ph.D.
Bonnie Weiss McLeod, Ph.D.
Cooley, LLP
mtuscan@cooley.com
bweissmcleod@cooley.com
zpatdocketing@cooley.com

BoxInterferences@uspto.gov
Telephone: 571-272-4683 Entered: September 3, 2014

UNITED STATES PATENT AND TRADEMARK OFFICE
________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
________________

Patent Interference No. 105,970 (JGN)

INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE
and Ti-Shiue Biotech, Inc.
Junior Party
(US 8,846,630)

v.

PACIFIC BIOSCIENCES
of California, Inc.
Senior Party
(13/633,673 and 13/930,178)
________________

Before SALLY GARDNER LANE, JAMES T. MOORE, and JOHN G. NEW,
Administrative Patent Judges.

NEW, Administrative Patent Judge.

DECISION ON MOTIONS
37 C.F.R. § 41.125

Interference 105,970

2

I. Introduction 1
A. Background 2
Interference 105,970 is before a panel of the Board for a decision on motions. 3
Oral argument was requested.
1
4
We have considered the record submitted by the parties. 5
B. Abbreviations 6
The following abbreviations are used in the record and this opinion. 7
Pacific Biosciences Motion 1 (PacBio Motion 1) 8
Pacific Biosciences Motion 2 (PacBio Motion 2) 9
Industrial Technology Research Institute Motion 1 (ITRI Motion 1) 10
Industrial Technology Research Institute Motion 2 (ITRI Motion 2) 11
Industrial Technology Research Institute Motion 3 (ITRI Motion 3) 12
Industrial Technology Research Institute Opposition 1 (ITRI Opp. 1) 13
Industrial Technology Research Institute Opposition 2 (ITRI Opp. 2) 14
Pacific Biosciences Opposition 1 (PacBio Opposition) 15
C. Parties 16
1. Junior Party 17
The Junior Party is Pacific BioSciences of California, Inc. (“PacBio”), which 18
is also the real party-in-interest.
2
19
20
2. Senior Party 21
The Senior Party is Industrial Technology Research Institute

1
Paper Nos. 155, 156
2
Paper No. 5
Interference 105,970

3

and Ti-Shiue Biotech, Inc. (“ITRI”), which are also the real parties-in-1
interest.
3
2
3
D. Subject matter involved in interferences 4
1. The counts 5
Count 1 is the sole count in this interference.
4
6
2. Count 1 7
ITRI claim 24: 8
9
A method of determining a sequence of a double-stranded nucleic acid 10
sample and a position of at least one modified base in the 11
sequence, comprising: 12
13
a. locking the forward and reverse strands of the nucleic acid sample 14
together to form a circular pair-locked molecule; 15
16
b. obtaining sequence data of the circular pair-locked molecule via 17
single molecule sequencing, wherein sequence data comprises 18
sequences of the forward and reverse strands of the circular 19
pair-locked molecule; and 20
21
c. determining the sequence of the double stranded nucleic acid 22
sample and the position of the at least one modified base in the 23
sequence of the double stranded nucleic acid sample by 24
comparing the sequences of the forward and reverse strands of 25
the circular pair-locked molecule, wherein at least one modified 26
base in the double stranded nucleic sample is paired with a base 27
having a base pairing specificity different from its preferred 28
partner base. 29
30
The claims of the parties are: 31

3
Paper No. 9, 17
4
Papers Nos. 1, 28
Interference 105,970

4

ITRI: Claim 1-28; all corresponding to Count 1 1
PacBio: 13/633,673 application; claim 1 corresponding to Count 1 2
13/930,178 application: claims 1-4, all corresponding to 3
Count 1
5
4
5
With respect to Count 1, the parties have been accorded the benefit of the 6
following applications: 7
ITRI: Provisional application 61/167,313, filed 7 April 2009

PacBio: Application 12/635,618, filed 10 December 2009
Provisional application 61/201,551, filed 11 December 2008
8
E. Subject matter involved 9
Knowing the sequence of the nucleotide base pairs in a molecule of nucleic acid, 10
such as DNA, provides the fundamental key to understanding genomic structure and 11
function. A molecule of DNA consists of two strands (5ˊ-3ˊ and 3ˊ-5ˊ or “forward” and 12
“reverse” strands) of paired nucleotide bases, the bases being paired according to the 13
Crick-Watson nucleotide base-pairing rules, i.e., adenine–thymine and cytosine–guanine. 14
The forward and reverse strands therefore usually contain complementary and redundant 15
information, meaning that the sequence of one strand can be deduced from the other 16
using the base pairing rules. 17
Chemical modifications to the nucleotide bases can result from exposure to 18
damaging agents such as radiation or reactive oxygen species and can have epigenetic or 19
mutagenic effects. ITRI Motion 2
6
at 3. It is therefore often important to know not only 20
the base sequence of a nucleic acid, but also the positions of modified bases. When 21

5
See Paper No. 28 at 3.
6
Paper No 59
Interference 105,970

5

Pacific Biosciences’ Appl. No. 61/201,551 (the “’551 application”) was filed, sequencing 1
technology could be used to obtain genetic sequence data from single nucleic acid 2
molecules using a sequencing-by-synthesis approach—a type of single molecule 3
sequencing. Id. at 3-4. Sequencing-by-synthesis involves determining the order in which 4
nucleotides are added during complementary nucleic acid strand synthesis, and then 5
determining the identity of the nucleotides in the original molecule using the usual base-6
pairing rules. Id. at 4. However, the results of sequencing-by-synthesis of a single DNA 7
strand based on the Crick-Watson rules do not identify whether or where chemical 8
modifications of the bases are present. Id. 9
Count 1 of this interference is claim 24 of ITRI’s US 8,846,630 B2
7
(the 10
“’630 patent”) (Ex. 2001) and is directed to detecting a mismatch in a constructed 11
circular pair-locked molecule (CPLM); a single circular molecule that contains within it 12
both the forward and reverse strands that are to be analyzed. Motion 1 at 4. CPLMs are 13
formed by ligating “hairpin” sequences to each end of the DNA segment to be analyzed, 14
one hairpin segment containing a primer sequence to initiate replication by DNA 15
polymerase. Performing sequencing-by-synthesis on a CPLM can then generate 16
sequence data for both of the original forward and reverse complementary strands. Id. 17
Disagreements in the analyzed sequences of the forward and reverse strands indicate that 18
the CPLM has a mismatched, modified base in at least one position; an investigator can 19
thus use the disagreement to determine the position of the modified base. Id. In Count 1, 20
one compares the sequences of the forward and reverse strands, and determines that a 21
modified base is present at positions where there is a disagreement. Id. at 5. 22
23

7
Paper No. 62
Interference 105,970

6

II. PacBio’s Motion 1
8
to rescind the benefit accorded ITRI for the 1
’3l3 application 2
3
PacBio moves to deny the benefit accorded to ITRI for its Provisional Application 4
61/167,313 (the “’3l3 application”) because it allegedly fails to provide a constructive 5
reduction to practice of Count 1. Motion at 1. According to PacBio, ITRI’s ’313 6
application fails to: (1) adequately provide a written description of an embodiment within 7
the scope of Count 1; and (2) enable an embodiment within the scope count 1. Motion at 8
2, 9. 9
A. Written Description 10
PacBio argues that ITRI’s ’313 application fails to provide an adequate written 11
description of two of the limitations of Count 1: (1) “locking the forward and reverse 12
strands of the nucleic acid sample together to form a circular pair-locked molecule”; and 13
(2) “obtaining sequence data of the circular pair-locked molecule by single molecule 14
sequencing.” Motion at 3, 8. 15
16
1. “locking the forward and reverse strands of the nucleic acid sample 17
together to form a circular pair-locked molecule” 18
19
PacBio argues that the ’313 application fails to describe the length of the double 20
stranded DNA sample that can be used in the manufacture of a CPLM. Motion at 3-4 21
(citing Ex. 1003
9
). Nor, argues PacBio, does the ’313 application provide a description 22
of the nucleotide length, base composition or the strandedness (e.g., double stranded, 23
single stranded or a combination) of the remaining components necessary to construct a 24

8
Paper No. 55
9
Paper No. 38
Interference 105,970

7

CPLM. Id. PacBio asserts that the series of steps required for constructing a CPLM and 1
the associated reaction conditions are similarly not provided in the ’313 application. Id. 2
PacBio contents that one of skill in the art would therefore be left to determine the nature 3
of the components, series of steps and reaction conditions necessary to build the CPLM 4
sequencing template: this, contends PacBio, does not meet the standard for achieving a 5
constructive reductive to practice. Id. (citing Goeddel v. Sugano, 617 F.3d 1350, 1355-6
1356 (Fed. Cir. 2010). 7
Furthermore, argues PacBio, Figures 1A and Figure 3A of the ’313 application 8
depict CPLMs that include gaps between the hairpin endpieces and the central double 9
stranded region such that the forward and reverse strands are not completely locked 10
together, and that no examples, prophetic or otherwise, are provided with respect to how 11
the forward and reverse strands of the nucleic acid sample are locked together. Motion at 12
4-5 (citing Ex. 1003). PacBio admits that although ligation reactions and hairpin DNA 13
sequences were publically available as of the ’313 application filing date, these features 14
were not mentioned in the ’313 application as useful for manufacturing a CPLM or 15
incorporated by reference in the application. Motion at 5 (citing Ex. 1003). 16
PacBio argues further that although the ’313 application recites “[t]he invention 17
discloses a method for preparing a unique DNA template by bridging the two 18
complementary strands of a DNA fragment together,” the application fails to describe 19
which steps must be performed to bridge the strands together to form the CPLM. Motion 20
at 5-6 (citing Ex. 1003). 21
Interference 105,970

8

PacBio adduces the Declaration of Dr. Gerald Zon
10
(the “Zon Declaration 1”) 1
(Ex. 1010). Dr. Zon holds a doctoral degree in organic chemistry from Princeton 2
University, has had a career of over 40 years in academia and the private sector, and has 3
authored over 270 publications. Id. Dr. Zon is currently Director of Business 4
Development for TriLink BioTechnologies. Id. After reviewing his curriculum vitae, we 5
find that Dr. Zon is an expert qualified to opine about the issues before the Board in this 6
interference. 7
Dr. Zon states that there are two classes of methods for constructing a CPLM: the 8
first is to create single stranded overhangs on a double stranded DNA fragment, followed 9
by hybridization of the double stranded DNA fragment to hairpin DNAs also having 10
single stranded overhangs, followed by gap filling and ligation to form the fully 11
circularized molecule. Motion at 6, (citing Ex. 1010, ¶¶14-15). The second is to create a 12
double stranded DNA fragment with blunt ends followed by blunt end ligation of hairpin 13
DNA to each side to circularize the molecule. Id. Zon declares that, with respect to the 14
first method, although the arrows in Figures 1A and 3A appear to depict a gap filling 15
step, there is no overhang shown to perform a hybridization step and that it is unclear 16
how one would fill the gaps. Id. (citing Ex. 1010, ¶ 16). 17
With respect to the second method, Zon declares that the arrows in Figures 1A and 18
3A of the ’313 application are not consistent with a blunt end ligation approach. Id. 19
(citing Ex. 1010, ¶ 17). 20
PacBio admits that, at the time the ’313 application was filed, PacBio’s own 21
disclosure of CPLM sequencing templates had been publically available for 22

10
Paper No. 96
Interference 105,970

9

approximately six months. Motion at 9 (Ex. 1009
11
at 15-22). However, argues PacBio, 1
Ex. 1009, which discloses how to assemble CPLMs for use as sequencing templates, was 2
neither incorporated-by-reference nor cited in the ’313 application. Id. 3
ITRI responds that PacBio has adduced no testimonial evidence with respect to 4
what the ’313 application would have conveyed to one of ordinary skill in the art, and 5
that the remainder of its arguments constitute mere attorney argument by those not 6
qualified to so opine. Opp. at 4. According to ITRI, in Zon Declaration 1 and also in 7
Zon Declaration 2
12
(Ex. 1020), Dr. Zon discusses Figures, 1A and 3A, but does not 8
discuss any of the text or the remaining figures in the application. Id. at 5 (citing 9
Ex. 1010, ¶¶ 1-19; 1020, ¶¶ 1-21). Moreover, argues ITRA, with respect to Figures 1A 10
and 3A, Dr. Zon says nothing about what those figures would have conveyed to one 11
skilled in the art. Id. Furthermore, ITRI argues, even if, arguendo, the Figures were 12
incompatible with making a CPLM, it would not meet PacBio’s burden of proof—ITRI 13
argues that PacBio must analyze the disclosure within the “four corners of the 14
specification” in light of the state of the art. Id. (quoting Ariad v. Lilly, 598 F.3d 1336, 15
1351 (Fed. Cir. 2010) (en banc)). 16
Consequently, argues ITRI, Dr. Zon’s testimony is fatally flawed because he does 17
not state whether the ’313 application would have conveyed possession of an 18
embodiment of the Count to one skilled in the art, despite declaring that he believed 19
himself to be able to comment on the level of ordinary skill in the art. Id. (citing Ex. 20
1010, ¶ 8). 21

11
Paper No. 44
12
Paper No. 103
Interference 105,970

10

Furthermore, argues ITRI, Dr. Zon declares that two methods involving hairpin 1
ligation could have been used to make a CPLM. Id. (citing Ex. 1010, ¶¶ 14-15). 2
Moreover, ITRI argues PacBio admits that hairpins and ligations were known. Id. (citing 3
Motion at 5, ll.11-12). Consequently, argues ITRI, Dr. Zon’s testimony, read together 4
with the ’313 application’s teaching to make a CPLM from target DNA, supports the 5
conclusion that the application would have conveyed possession of making a CPLM to 6
one of ordinary skill. Id. 7
Similarly, ITRI argues that PacBio provides no other evidence supporting its 8
argument that the ’313 application would not have conveyed, to one of ordinary skill in 9
the art, obtaining sequence data of a CPLM via single molecule sequencing. Opp. at 5. 10
ITRI points out that PacBio admits that “single molecule sequencing” appears in the 11
’313 application, and admits that methods of single molecule sequencing of a CPLM 12
were known. Id. (citing Motion at 8, ll. 15-16; 10, ll. 9-12; Ex. 1009 at 15-22). ITRI 13
asserts that a single mention of a term suffices for written description of well-known 14
subject matter. Id. (citing EnOcean GmbH v. Face Int’l Corp., 742 F.3d 955, 961 (Fed. 15
Cir. 2014). ITRI contends that PacBio’s pointing out that a term appears once cannot, by 16
itself, establish a lack of description. Opp. at 6. 17
ITRI maintains that the ’313 application describes and illustrates target DNA from 18
which the CPLM is made and that that a CPLM is made by connecting the strands of the 19
target DNA, and it showed that linkers are used to do so. Opp. at 6-7 (citing Ex. 2002 20
7 at 5, Fig. 2, ¶¶ 1-2; Ex. 2034, ¶¶ 35-37). It explained “that a CPLM is made by 21
connecting the strands of the target DNA, and it showed that linkers are used to do so.” 22
Opp. at 7. (Ex. 2002, p. 5, paragraphs 3-4; p. 6, Fig. 4A.1; p. 7, Fig. A2; Ex. 2034, ¶ 37). 23
ITRI contends that a person of ordinary skill would have understood from the ’313 24
Interference 105,970

11

application that the inventors described a circular DNA molecule having a double-1
stranded “target DNA” region bracketed by two linkers that connect the two strands and 2
that one linker in the molecule contains a binding site for a sequencing primer. Opp. at 7 3
(citing Ex. 2034, ¶¶ 37, 42). 4
In support of this contention, ITRI adduces the declaration of Dr. Shawn Levy 5
(Ex. 2034, the “Levy declaration”). Dr. Levy received his doctoral degree in 6
Biochemistry and Genetics from Emory University in 2000 and is currently a Faculty 7
Investigator and the Director of the Genomic Services Laboratory at the HudsonAlpha 8
Institute for Biotechnology in Huntsville, Alabama. Id. at ¶¶ 1, 4. Dr. Levy has 9
published more than 90 peer-reviewed articles on the subject of the application of 10
genomic technologies. Id. at ¶ 6. Having reviewed Dr. Levy’s Declaration we find that 11
he is an expert qualified to opine on the issues before the Board in the instant 12
interference. 13
Dr. Levy states that one of ordinary skill in the art would have a Ph.D. in molecular 14
biology or a related field and two years’ experience in a research laboratory where 15
sequencing is commonly performed. Ex. 2034, ¶ 15. Dr. Levy declares that one so 16
skilled in the art would have known of at least three ways to attach the hairpin linkers, 17
each involving ligation and that such ligation reactions were well known in the art as of 18
April 2009. Ex. 2034, ¶ 51. 19
Consequently, Dr. Levy attests, the design of ligation reactions that would convert 20
a target DNA to a CPLM, e.g., by attaching hairpin sequences would have been within 21
the level of ordinary skill as of April 7, 2009, given the guidance in the ’313 application 22
Ex. 2034, ¶¶ 19-20 (citing Ex. 2002 at Fig. 2, Fig. 4A.1 and accompanying text). Dr. 23
Levy also declares that the outcome of these ligation reactions would have been 24
Interference 105,970

12

predictable to one skilled in the art based on both general understanding of ligation 1
reactions and specific, known examples of hairpin ligations. Id. 2
ITRI also adduces Exhibits 2037, 1009, and 2038 as examples demonstrating that 3
ligation techniques were well known and predictable in the art at the time the 4
’313 application was filed. 5
6
2. “obtaining sequence data of the circular pair-locked molecule by single 7
molecule sequencing” 8
9
PacBio concedes that the ’313 application discusses the use of “single molecule 10
sequencing” but argues that it includes only a single mention that the technique can be 11
used in conjunction with a pair locked molecule. Motion at 9 (citing Ex. 1003 at 3). 12
According to PacBio, the ’313 application is silent regarding the steps necessary to carry 13
out a single molecule sequencing reaction on a CPLM, the types of sequencing reactions 14
that are amenable for use with a CPLM, and whether the CPLM is assembled in a 15
different manner based on the sequencing reaction employed. Motion at 9 (citing Ex. 16
1003). 17
ITRI responds that single molecule sequencing techniques were well known and 18
points to PacBio’s US 7,302,146, (Ex. 2036
13
, the “’146 patent”) issued in 2007, that 19
discloses single molecule sequencing of circular templates. Opp. at 11 (citing, e.g., 20
Exs. 2036; 2034, ¶¶ 45-46). Dr. Levy declares that the ’146 patent teaches that its 21
“sequencing methods can be used” with “any nucleic acid molecule, including double-22
stranded or single-stranded, linear or circular nucleic acids (e.g., circular DNA), single 23
stranded DNA hairpins,” and various other structures including “complex nucleic acid 24

13
Paper No. 113
Interference 105,970

13

structures” such as “tandem repeats.” Id. (citing Ex. 2034, ¶ 45; Ex. 2036, col. 23, ll. 50-1
58). Therefore, Dr. Levy declares, one skilled in the art would have known of a single 2
molecule sequencing method suitable for sequencing a circular nucleic acid molecule, 3
such as a circular pair-locked nucleic acid molecule. Id. (citing Ex. 2034, ¶ 45). 4
Furthermore, there would have been no concern that double-stranded regions or hairpin 5
linkers contained within a CPLM might present difficulties because the ’146 patent 6
expressly indicates that its methods work with all of these types of structures. Id. 7
8
3. The ’313 application conveyed possession of a single embodiment within 9
count 1 10
11
We are not persuaded by either of PacBio’s arguments. 12
As an initial matter, Dr. Zon opines that, as of April 7, 2009, the ordinarily skilled 13
artisan had an advanced degree (i.e., M.S. or Ph.D. degree) in molecular biology, 14
biochemistry, biophysics, bioengineering, or related field, and one or more years post-15
graduate experience. Ex. 1010, ¶ 8. Similarly, Dr. Levy opines that, at approximately the 16
same period, one of ordinary skill in the art of single molecule sequencing and detection 17
of modified bases would have a Ph.D. in molecular biology or a related field and two 18
years’ experience in a research laboratory where sequencing is commonly performed. 19
Ex. 1034, ¶ 15. We agree with the parties that a person of ordinary skill in the art would 20
have possessed a Ph.D. in molecular biology or a related field and 1-2 years’ postdoctoral 21
experience in a position where sequencing is routinely performed. Having set the 22
standard of ordinary skill, therefore, at a relatively high level, we assume that a person of 23
ordinary skill would have a sophisticated understanding of the molecular biology of 24
nucleic acids and their synthesis. 25
Interference 105,970

14

Benefit of an earlier-filed application in an interference requires that the 1
application would have conveyed possession of a single embodiment within the Count to 2
one of ordinary skill in the art. Bd.R. 201; see also Hunt v. Treppschuh, 523 F.2d 1386, 3
1389 (C.C.P.A. 1975). The only substantive testimonial evidence that PacBio adduces in 4
support of its arguments is the Zon Declarations. Dr. Zon declares that: 5
Based on my understanding of the invention claimed in the 6
’630 patent and the nature of a circular pair-locked molecule (cPLM), 7
I envision two classes of methodology for constructing a cPLM. 8
9
The first is to create single stranded overhangs on a double stranded 10
DNA fragment, followed by hybridization of the double stranded DNA 11
fragment to hairpin DNAs also having single stranded overhangs, followed 12
by gap filling and ligation to form the fully circularized molecule. The 13
second is to create a double stranded DNA fragment with blunt ends 14
followed by blunt end ligation of hairpin DNA to each side to circularize the 15
molecule. 16
17
With respect to the first method, although the arrows in Figures 1A 18
and 3A appear to depict a gap filling step, there is no overhang shown to 19
perform a hybridization step. Additionally, it is unclear how one would fill 20
the gaps identified by the red circles in Figures 1 A and 3A. 21
22
With respect to the second method, the arrows in Figures 1A and 3A 23
are not consistent with a blunt end ligation approach. 24
25
Accordingly, Figures 1A and 3A are not compatible with the steps 26
and/or components necessary to construct a cPLM. 27
28
Ex. 1010, ¶¶ 14-18. The entire relevant portion of the Zon Declaration, therefore, deals 29
with whether, in Dr. Zon’s opinion, Figures 1A and 3A are “compatible with the steps 30
and/or components necessary to construct a CPLM.” Ex. 1010, ¶ 18. 31
Interference 105,970

15

However, Zon does not discuss the state of the prior art, other than to observe that 1
it was well known in the art to store a double-stranded DNA sequencing template at or 2
below 25
o
C and in the absence of divalent cations. Ex. 1010, ¶¶ 11, 12. Dr. Zon is silent, 3
however, with respect to the known state of the art of ligation and the construction of 4
CPLMs at the time the ’313 application was filed. Nor does Dr. Zon declare what a 5
person of ordinary skill at the time of invention would have understood of the 6
construction of CPLMs or whether such a skilled artisan would be in possession of the 7
invention. And Dr. Zon is also silent with respect to the subject of sequencing methods, 8
single molecule or otherwise. 9
By contrast, ITRI’s declarant, Dr. Levy, states that: 10
[O]ne skilled in the art would have known of at least three ways to attach the 11
hairpin linkers. Each of these involved ligation. Ligation reactions were well 12
known in the art. They had been known and performed for decades as of 13
April 2009. And ligations were a common, basic step in molecular biology 14
cloning techniques … Furthermore, the outcome of these ligation reactions 15
would have been predictable to one skilled in the art based on both general 16
understanding of ligation reactions and specific, known examples of hairpin 17
ligations. 18
19
Thus, there was knowledge in the art of relevant procedures for 20
making a CPLM, and these procedures were predictable. One of ordinary 21
skill would have understood which procedures would have been suitable 22
given the guidance in the ’313 application. Accordingly, there would have 23
been little, if any, need for one of ordinary skill in the art to experiment to 24
lock the forward and reverse strands of a nucleic acid sample together to 25
form a circular pair-locked molecule. 26
27
Ex. 2034, ¶¶ 51, 52 (internal citations omitted). Moreover, Dr. Levy declares that: 28
Techniques for single molecule sequencing of various types of DNA, 29
including circular templates, were known. A CPLM is circular and contains 30
Interference 105,970

16

hairpins and single- and double-stranded regions. Pacific Biosciences’ 1
’146 patent disclosed methods of “Single Molecule Sequencing” and 2
expressly stated that its “sequencing methods can be used” with “any nucleic 3
acid molecule, including double-stranded or single-stranded, linear or 4
circular nucleic acids (e.g., circular DNA), single stranded DNA hairpins,” 5
and various other structures including “complex nucleic acid structures” 6
such as “tandem repeats.” 7
… 8
Thus, for a CPLM made by attaching hairpin linker oligonucleotides to 9
target DNA of the length produced by nebulization (700-1330 bp), one 10
skilled in the art would predictably have been able to obtain forward and 11
reverse strand sequences from a single molecule sequencing reaction. 12
13
Ex. 2034, ¶ 53. (citations omitted). 14
In addition, ITRI adduces documentary evidence of the state of the art at the time 15
of the ’313 application filing. ITRI’s Exhibit 2044
14
is an article entitled The 3ˊ to 5ˊ 16
exonuclease activity of Mre11 facilitates repair of DNA double-strand breaks by Tanya 17
T. Pauli and Martin Gellert and published in the journal MOLECULAR CELL, Vol. 1, 969–18
979, in June 1998. The article states, inter alia, that: “[T]he dumbbell substrate was 19
composed of two [DNA] hairpins ligated together.” Ex. 2044 at 978. 20
Similarly, ITRI’s Exhibit 2037
15
is a European Patent Application, EP 0 653 489 21
A1, entitled Coccidiosis vaccin [sic], dated September 9, 1994. The application 22
discloses: “The remaining cDNA was end-repaired using T4 DNA polymerase in the 23
presence of all four dNTPs at 37
o
C for 30 minutes. EcoRI adaptors were ligated onto the 24
blunted ends of the cDNA using T4 DNA ligase at 8
o
C for 24 hours” as an example of 25
blunt-end ligation techniques known in the art. Ex. 2037, p. 11, ll. 24-26. 26

14
Paper No. 121
15
Paper No. 114
Interference 105,970

17

ITRI also points to its Exhibit 2038
16
, entitled Accurate whole human genome 1
sequencing using reversible terminator chemistry, by David R. Bentley et al., which was 2
published in NATURE, vol. 456, pp. 53-59 on November 6, 2008. On pages 1-2 of the 3
supplemental information to the article (published by NATURE), the article teaches 4
fragmentation of purified DNA by nebulization and subsequent ligation using a single-A 5
overhang. Ex. 2038 at 1-2. 6
Similarly, with respect “obtaining sequence data …by single molecule 7
sequencing,” PacBio adduces no substantial evidence in support of its argument that a 8
person of ordinary skill in the art would not have known that ITRI was in possession of 9
the invention disclosed in the ’313 application. As noted supra, the Zon Declarations are 10
silent with respect to the subject. 11
In contrast, ITRI points to PacBio’s ’146 patent, published November 7, 2007, 12
entitled Apparatus and Method for Analysis of Molecules, which discloses “single 13
molecule sequencing,” which: 14
[C]an be used to determine the nucleic acid of any nucleic acid molecule, 15
including double-stranded or single-stranded, linear or circular nucleic acids 16
(e.g., circular DNA), single stranded DNA hairpins, DNA/RNA hybrids, 17
RNA with a recognition site for binding of the polymerase, or RNA hairpins. 18
The methods of the present invention are suitable for sequencing complex 19
nucleic acid structures, such as 5' or 3' non-translation sequences, tandem 20
repeats, exons or introns, chromosomal segments, whole chromosomes or 21
genomes. 22
23
Ex. 2036, col. 23, ll. 30; 50-59. Dr. Levy declares that, in addition to the ’146 patent, 24
multiple publications had discussed single molecule sequencing, including John Eid et al. 25
Real-time DNA sequencing from single polymerase molecules, 323 SCIENCE 133-38 26

16
Paper No. 115
Interference 105,970

18

(January 2, 2009) (Ex. 2039
17
), Pushpendra K. Gupta, Single-molecule DNA sequencing 1
technologies for future genomics research, 26(11) TRENDS IN BIOTECHNOLOGY 602-11 2
(August 21, 2008) (Ex. 2040
18
), Elaine R. Mardis, Next-Generation DNA Sequencing 3
Methods, 9 Ann. Rev. Genomics Hum. Genet. 387-402 (2008) (Ex. 2042
19
), and Jay 4
Shendure & Hanlee Ji, Next-generation DNA sequencing, 26(10) NATURE 5
BIOTECHNOLOGY 1135-1145 (October 2008) (Ex. 2046
20
), which teaches: 6
[A] highly sensitive fluorescence detection system is used to directly 7
interrogate single DNA molecules via sequencing by synthesis. Template 8
libraries, prepared by random fragmentation and poly-A tailing (that is, no 9
PCR amplification), are captured by hybridization to surface-tethered poly-T 10
oligomers to yield a disordered array of primed single-molecule sequencing 11
templates. At each cycle, DNA polymerase and a single species of 12
fluorescently labeled nucleotide are added, resulting in template-dependent 13
extension of the surface-immobilized primer-template duplexes. 14
15
Ex. 2046 at 1140 (emphasis added); Ex. 2034, ¶ 46. 16
Thus, on the record before us, we determine that ITRI has not shown that the ’313 17
application lacks description of an embodiment of the Count when the ’313 application is 18
considered from the perspective of a person of ordinary skill in the art. 19
20
B. Lack of Enablement 21
PacBio next argues that the ’313 application fails to include a working example 22
pertaining to the subject matter of the count, nor does it include any working examples 23
related to the individual method steps set forth in the count. Motion at 10. According to 24

17
Paper No. 116
18
Paper No. 117
19
Paper No. 119
20
Paper No. 123
Interference 105,970

19

PacBio, there are, for instance, no working examples pertaining to construction of a 1
CPLM; specifically locking the forward and reverse strands of a nucleic acid sample 2
together to form a circular pair-locked molecule. Id. Nor, asserts PacBio, does the 3
’313 application include a working example directed to obtaining sequence data of the 4
CPLM by single molecule sequencing. Motion at 11. 5
PacBio asserts that, where the claimed invention is the application of an 6
unpredictable technology in the early stages of development, an enabling description 7
must provide a specific and useful teaching. Motion at 11 (citing Genentech, Inc. v. Novo 8
Nordisk, AIS, 108 F.3d 1361, l367-l368 (Fed. Cir. 1997); Chiron Corp. v. Genentech, 9
Inc., 363 F.3d 1247, 1254 (Fed. Cir. 2004). PacBio alleges that the ’313 application fails 10
to provide any such teaching. Id. 11
ITRI responds that the ’313 application provided substantial guidance for making a 12
CPLM by locking forward and reverse strands of a double stranded nucleic acid together. 13
First, argues ITRI, it teaches that a “DNA molecule with high molecular weight is cut 14
into smaller fragments (target DNAs)”; more specifically, “[g]enomic DNA are shredded 15
to smaller fragments (target DNAs)” which are double-stranded, i.e., have forward and 16
reverse strands. Opp. at 17 (citing Ex. 2002, p. 5, ¶¶ 1-2, Fig. 2). According to ITRI, one 17
skilled in the art would have understood that a target DNA (a shredded piece of genomic 18
DNA) could be prepared using a well-known fragmentation technique, e.g., nebulization. 19
Id. (citing Ex. 2034, ¶¶ 36, 50; e.g., Ex. 2038, at ¶¶ 1-2; 2045, Abstract). 20
ITRI argues further that the ’313 application then teaches that “[e]ach target DNA 21
is connected together into a circular structure through a series of steps …. Each target 22
DNA is built into a pair-locked molecule construct, where both complementary strands of 23
this target DNA are connected.” Opp. at 17 (quoting Ex. 2002 at 5, ¶¶ 3-4). ITRI 24
Interference 105,970

20

repeats its argument supra that there were at least three known molecular biology 1
techniques for achieving this. Id. (citing Ex. 2034, ¶¶ 39-42, 51; Ex. 2044, at 978; 2
Ex. 2037, p. 11, ll. 22-26; Ex. 1009, at 30-31; Ex. 2038, ¶¶ 1-2 of SI; Motion 1, p. 5, 3
ll. 11-12. ITRI contends, therefore, that the level of experimentation needed to perform 4
these procedures to make a CPLM and the unpredictability associated with them would 5
have been minimal. Opp. at 17. 6
ITRI argues further that, with respect to single-molecule sequencing, PacBio had 7
already enabled that invention in its ’146 patent. Opp. at 18. Furthermore, contends 8
ITRI, a CPLM has no features that would require substantial experimentation to adapt the 9
technique in PBC’s ’146 patent to a CPLM. Opp. at 18-19 (citing Ex. 2034, ¶¶ 53-54). 10
On the contrary, ITRI argues, one skilled in the art would have correctly predicted that 11
the ’146 technique would work with a CPLM. Id. 12
The standard for enablement under 35 U.S.C. § 112 (first paragraph) is whether 13
one of ordinary skill in the art could have made and used the invention without undue 14
experimentation. See In re Wands, 858 F.2d 731, 737 (Fed. Cir. 1988). In the context of 15
an interference, enablement is required only for one embodiment within the Count. Hunt, 16
523 F.2d at 1389. Furthermore, the question of enablement is whether the disclosure is 17
sufficient to enable those skilled in the art to practice the claimed invention; therefore, the 18
specification need not disclose what is well known in the art. In re Myers, 410 F.2d 420, 19
(CCPA 1969). 20
PacBio has adduced no persuasive evidence or testimony to show that a person of 21
ordinary skill in the art could not have made and used the invention without undue 22
experimentation. As ITRI has demonstrated supra, the limitations of the count that 23
PacBio is contesting were available in the prior art prior to the filing of the ’313 patent. 24
Interference 105,970

21

For example, Dr. Levy, ITRI’s declarant states that: 1
One of ordinary skill would have understood which procedures would have 2
been suitable given the guidance in the ’313 application. Accordingly, there 3
would have been little, if any, need for one of ordinary skill in the art to 4
experiment to lock the forward and reverse strands of a nucleic acid sample 5
together to form a circular pair-locked molecule” 6
7
and that: “One skilled in the art would also have known how to use well known, 8
predictable techniques for single molecule sequencing to obtain sequence data for 9
the forward and reverse strands of the CPLM in accordance with the 10
’313 application disclosure with little, if any, need for experimentation.” Ex. 2034, 11
¶¶ 52, 54. Moreover, Dr. Levy supports this conclusion by citing substantial 12
evidence; against which PacBio adduces no persuasive evidence or testimony to 13
rebut. See, e.g., Ex. 2034, ¶¶ 36-52. We therefore conclude that PacBio has not 14
met its burden of showing that an embodiment of the Count is not enabled by the 15
‘313 disclosure. PacBio’s Motion 1 is consequently DENIED. 16
17
III. PacBio Motion 2
21
for judgment of unpatentability under 35 U.S.C. § 103 18
PacBio’s Motion 2 seeks judgment against ITRI on the ground that all of the 19
involved claims 1-28 of ITRI’s US 8,486,630 (the “’630 patent”)
22
would have been 20
obvious in view of prior art under 35 U.S.C. § 103(a). Motion at 1. Specifically, PacBio 21
alleges that the involved claims would have been obvious over the combination of 22

21
Paper No. 56
22
Paper No. 2001
Interference 105,970

22

PacBio’s US 8,153,375
23
(filed March 27, 2009) (Ex. 1011
24
) in combination with C.D. 1
Laird et al., Hairpin-bisulfite PCR: Assessing epigenetic methylation patterns on 2
complementary strands of individual DNA molecules, 101(1) PROC. NAT. ACAD. SCI. 204. 3
(January 6, 2004) (Ex. 1012
25
) (“Laird”) or T. Matsumura et al., Photochemical transition 4
of 5-methylcytosine to thymine by DNA photoligation, NUCLEIC ACIDS SYMPOSIUM 5
SERIES NO. 51, 233 (Oxford Univ. Press, 2007) (Ex. 1013
26
) (“Matsumura”).
27
6
PacBio argues that independent claims 1, 24, and 26 are all based on the use of a 7
CPLM to obtain the sequence of the forward and reverse strands of a double stranded 8
nucleic acid, and thereafter to determine the positions of modified bases in the double-9
stranded nucleic acid. Motion at 4 (citing Ex. 1002). Claim 1 of the ’630 patent is 10
exemplary: 11

23
(The ’375 patent was filed March 27, 2009 and issued April 10, 2012). The application
which issued as the ’375 patent, US 2009/0298075 Al, was published on Dec. 3, 2009.
The ’375 patent claims priority to provisional applications 61/099,696, filed
September 24, 2008 (Ex. 1006), and 61/072,160, filed March 28, 2008 (Ex. 1007)
(respectively, “the ’696 application” and “the ’160 application”).
24
Paper No. 46
25
Paper No. 47
26
Paper No. 48
27
PacBio also argues that, if PacBio’s Motion 1 were granted, then ITRI should receive
only the benefit of the filing date of the ’630 patent, which is November 5, 2009. Motion
at 3. PacBio argues that, if it should prevail on Motion 1, PacBio’s Cold Spring Harbor
Personal Genomes Meeting Presentation (October 12, 2008) (Ex. 1009) (Paper No. 44)
(“the Personal Genomes presentation”) would qualify as a prior art reference under
35 U.S.C. § 102(b). Id. As we have related supra, PacBio’s Motion 1 is denied because
we determined that PacBio did not show that the ITRI benefit application lacked a
constructive reduction to practice of the Count. Under the circumstances before us, and
as PacBio has not advanced an argument that the Personal Genomes Meeting
Presentation is prior art if Motion 1 is denied, we have not considered it further in our
decision.
Interference 105,970

23

1. A method of determining a sequence of a double-stranded nucleic acid 1
sample and a position of at least one modified base in the sequence, 2
comprising: 3
4
a. locking the forward and reverse strands together to form a circular pair-5
locked molecule; 6
7
b. obtaining sequence data of the circular pair-locked molecule via single 8
molecule sequencing, wherein the sequence data comprises sequences of 9
the forward and reverse strands of the circular pair-locked molecule; 10
11
c. determining the sequence of the double-stranded nucleic acid sample by 12
comparing the sequences of the forward and reverse strands of the circular 13
pair-locked molecule; 14
15
d. altering the base-pairing specificity of bases of a specific type in the 16
circular pair-locked molecule to produce an altered circular pair-locked 17
molecule; 18
19
e. obtaining the sequence data of the altered circular pair-locked molecule 20
wherein the sequence data comprises sequences of the altered forward and 21
reverse strands; and 22
23
f. determining the positions of modified bases in the sequence of the double-24
stranded nucleic acid sample by comparing the sequences of the altered 25
forward and reverse strands. 26
27
Ex. 2001. The ’630 patent teaches a CPLM constructed by ligating hairpin inserts to both 28
ends of a double-stranded nucleic acid. Motion at 4. 29
PacBio asserts that, more than a year prior to the benefit currently accorded to 30
ITRI, PacBio filed the ’160 application, which teaches ligating hairpin adaptors to the 31
ends of a double-stranded DNA fragment to generate a CPLM (the “SMRTbell”™ 32
template”), allegedly intended for use as a template for single molecule real-time 33
Interference 105,970

24

(“SMRT”) sequencing. Motion at 4-5. According to PacBio, a SMRTbell™ has the 1
same structure and components as a CPLM in that it contains a double-stranded region 2
flanked by two hairpins and, further, that polymerase extension from a primed 3
SMRTbell™ template, as for a CPLM, generates sequence information for both the 4
forward and reverse strands. Motion at 5. 5
PacBio contends that although the ’375 patent discloses the use of a CPLM for 6
sequencing the forward and reverse strands of a single molecule double-stranded nucleic 7
acid template, neither explicitly discloses the use of the template to determine the 8
position of a modified base, or altering the base pairing specificity of a specific type of 9
base in the template and thereafter determining the position of the modified base, as 10
recited in ITRI’s claims 1, 24 and 26. Motion at 5. However, argues PacBio, ITRI’s 11
’630 patent acknowledges that such methods were known in the art. Id. (citing Ex. 1002, 12
cols. 22-24). Specifically, the ’630 patent describes Laird’s method of converting 13
cytosine (but not 5-methylcytosine) to uracil via bisulfite treatment and Matsumura’s 14
photochemical conversion of 5-methylcytosine (but not unmodified cytosine) to thymine. 15
Id. at 6 (citing Ex. 2002, col. 23, ll. 57-65; col. 23, ll. 49-56). PacBio asserts that both 16
base conversion methods result in a modified base that is paired with a base having a 17
base-pairing specificity different from the preferred partner base of the modified base.
28
18
PacBio argues that it would have been obvious at the time that ITRI’s 19
’313 application was filed (April 7, 2009) to construct and determine the sequence of a 20
double-stranded nucleic acid in a CPLM as taught by the ’375 patent and to subject the 21
CPLM to either (i) bisulfite conversion or (ii) photochemical conversion to create a 22

28
For example, a cytosine converted to uracil would be found to be base-paired with
guanine rather than the expected adenine.
Interference 105,970

25

modified base by which the position of a 5-methylcytosine residue could be determined 1
by comparing the sequences of the forward and reverse strands, as recited in claims 1, 5, 2
6, and 24-26 of ITRI’s ’630 patent. Motion at 6. 3
According to PacBio, one of ordinary skill in the art would have been motivated to 4
compare the sequences of the forward and reverse strands because the ’375 patent 5
explicitly teaches that comparing the sequence of the forward and reverse strand is one of 6
the advantages of using a CPLM, e.g., to provide redundant sequence information. Id. 7
PacBio argues that one of ordinary skill would have been further motivated to compare 8
the forward and reverse strands in order to increase the accuracy of 5-methylcytosine 9
determination, given the well-known importance of DNA methylation in the regulation of 10
gene expression, genomic imprinting and X-chromosome inactivation. Id. at 6-7 (citing 11
Ex. 1002, col. 2, ll.14-23). 12
Finally, argues PacBio, an artisan of ordinary skill would have had a reasonable 13
expectation of success in determining the location of a modified base in a CPLM, given 14
that the techniques for determining the location of 5-methylcytosine residues were 15
known, and the claimed methods simply involve adding or employing the unique CPLM 16
template in this known sequencing reaction. Id. at 7 (citing Ex. 1002, col. 23, ll.49-65, 17
Abstract; Ex. 1013, Abstract). PacBio contends that because the CPLM sequencing 18
template, single molecule sequencing and base conversion of a sequencing template had 19
previously been disclosed in the prior art, ITRI's invention amounts to a simple 20
combination of prior art elements according to known methods to achieve predictable 21
results. 22
PacBio further contends that claims 2-4 are and 7-23 are all dependent on claim 1, 23
and incorporate all of its limitations. Motion at 8. Moreover, argues PacBio, the 24
Interference 105,970

26

additional limitations recited in the dependent claims are also found in the ’375 patent 1
and therefore do not provide a basis for patentability.
29
2
PacBio also argues that claims 27-28 are also based on the use of a CPLM to 3
obtain the sequence of the forward and reverse strands of a double-stranded nucleic acid, 4
and thereafter to determine the positions of modified bases in the double-stranded nucleic 5
acid as discussed for the other claims above. Motion at 13. Claim 27 is exemplary and 6
recites: 7
27. A method of determining a sequence of a double-stranded nucleic 8
acid sample and a position of at least one modified base in the 9
sequence, comprising: 10
11
a. locking the forward and reverse strands together to form a circular 12
pair-locked molecule; 13
14
b. obtaining sequence data of the circular pair-locked molecule via 15
single molecule sequencing, wherein the sequence data comprises 16
sequences of the forward and reverse strands of the circular pair-17
locked molecule; 18
19
c. determining the sequence of the double-stranded nucleic acid 20
sample by comparing the sequences of the forward and reverse 21
strands of the circular pair-locked molecule· 22
23
d. obtaining sequencing data of the circular pair-locked molecule via 24
single molecule sequencing, wherein at least one nucleotide 25
analog that discriminates between a base and its modified form is 26
used to obtain sequence data comprising at least one position 27
wherein the at least one differentially labeled nucleotide analog 28
was incorporated; and 29

29
PacBio argues separately with respect to claims 2, 4, 7-10, 13, 14, 17-23, and 27-28.
Motion at 8-13. ITRI does not respond to these individual arguments in its opposition.
Interference 105,970

27

1
e. determining the positions of modified bases in the sequence of the 2
double-stranded nucleic acid sample by comparing the sequences 3
of the forward and reverse strands. 4
5
Ex. 2001. PacBio points out that, rather than employ a difference in base pairing 6
specificity (as in claims 1-26), claims 27 and 28 refer to the use of nucleotide analogs that 7
discriminate between a base and its modified form. Motion at 13. 8
According to PacBio, the only difference between these two independent claims is 9
the number of steps in the sequence determination, i.e., claim 27 employs a preliminary 10
sequencing step prior to sequencing using discriminating bases, whereas claims 28 does 11
not. Id. Furthermore, argues PacBio, although ITRI was accorded benefit of the ’313 12
application, the subject matter for claims 27 and 28 was not included in the application, 13
and was added only in ITRI’s U.S. Application No. 12/613,291 (the application that 14
matured into the ’630 patent). Id. (citing Exs. 1002; 1003). 15
PacBio asserts that the ’375 patent teaches the use of a CPLM for single molecule 16
sequencing, which ITRI admits in its ’313 priority application is the key point of novelty. 17
Motion at 13-14 (citing Ex. 1003, p. 3). Otherwise, contends PacBio the sequencing 18
techniques disclosed and claimed in the ’630 patent were previously known to one of 19
ordinary skill in the art: in particular, the use of discriminating nucleotide analogs to 20
identify modified bases, including 5-methylcytosine bases, was disclosed in 21
US 7,399,614 patent (the “’614 patent”), as acknowledged in the ’630 patent. Id. at 14 22
(citing Ex. 1002, col. 24, ll. 42-46; Ex. 1014
30
, Abstract, claims 1-29). 23

30
Paper No. 49
Interference 105,970

28

According to PacBio, it would have been obvious at the time that the ’630 patent 1
was filed to construct and determine the sequence of a double-stranded nucleic acid in a 2
CPLM, as disclosed in the ’375 patent, and to include discriminating nucleotide analogs 3
in the sequencing reaction as disclosed by the ’614 patent, to identify the position of 4
modified bases as recited in claims 27-28. Motion at 14. PacBio asserts that one of 5
ordinary skill would have been motivated to do so to increase the accuracy of the 6
sequence determination, given the importance of DNA methylation in the regulation of 7
gene expression, genomic imprinting, and X-chromosome inactivation. Id. Further, 8
PacBio contends that the ordinarily skilled artisan would have had a reasonable 9
expectation of success, given that discriminating analogs for determining the location of 10
5-methylcytosine residues were known, and the claimed methods simply involve 11
employing the unique CPLM template. Id. 12
ITRI responds that independent claims 1, 24, and 26 require using a disagreement 13
between the forward and reverse strand sequences of the CPLM to determine a modified 14
base’s position. Opp. at 2. According to ITRI, claim 24, which it chooses as an 15
exemplar, recites determining the position of at least one modified base in the sequence 16
of a nucleic acid sample “by comparing the sequences of the forward and reverse strands” 17
and that at least one modified base in the double stranded nucleic acid sample “is paired 18
with a base having a base pairing specificity different from its preferred partner base,” 19
which is referred to as being “mismatched.” Id. at 2-3. ITRI argues that claim 24 thus 20
teaches sequencing the forward and reverse strands of a CPLM containing a mismatched 21
modified base, such as a uracil paired to guanine instead of adenine, and that obtaining 22
these sequences reveals a disagreement which is used to conclude that a modified base 23
Interference 105,970

29

was present at the position of the disagreement. Id. at 3 (citing, e.g., Ex. 2048
31
, ¶¶ 31-1
33). 2
ITRI contends that it would be unreasonable to interpret claim 24 as covering a 3
determination of the position of the modified base that does not use the disagreement 4
between the two strands of DNA, because the claim requires a mismatched modified base 5
in the double stranded nucleic acid sample. Opp. at 4 (citing Ex. 2048, ¶ 34). According 6
to ITRI’s declarant, Dr. Levy: 7
One of ordinary skill would not understand claim 24 as meaning that the 8
disagreement between the sequence data of the forward and reverse strands 9
can be disregarded or that a modified base position can be determined using 10
only some other, undefined information independent of the disagreement. 11
Interpreting claim 24 to mean that a modified base position can be 12
determined using only some other, undefined information is unreasonable 13
because it would render the mismatch irrelevant and superfluous to the 14
determining step in which it appears. 15
16
Ex. 2048, ¶ 34. ITRI thus argues that it would not be reasonable to disregard the 17
disagreement, treat the mismatch as superfluous and irrelevant to the step that recites it, 18
and use a different basis—obtained by comparing the strands in some way not defined by 19
claim 24—to deduce a modified base position. Opp. at 4. 20
ITRI contends that PacBio provides no evidence supporting the attorney argument 21
that its proposed combination of documents would have rendered ITRI’s claims obvious 22
to one of ordinary skill in the art. Opp. at 7. ITRI points out that the Zon Declaration 23
takes no position concerning the alleged obviousness of claims 1, 24, and 26. Id. (citing 24
Ex. 1010, ¶¶ 9-18). 25

31
Paper No. 134
Interference 105,970

30

ITRI also argues that PacBio’s arguments fail to combine the teachings of its cited 1
prior art references in a manner that shows all of the claimed steps, including determining 2
the position of the modified base using a disagreement between the forward and reverse 3
strand sequences of the CPLM. Opp. at 7. ITRI points out that PacBio has failed to 4
submit a claim chart comparing the claim language to the prior art in conformance with 5
Bd.R. 41.121(e), as well as to construe the claims. ITRI argues that PacBio also fails to 6
explain why one of ordinary skill in the art would have modified the methods of its 7
’375 patent in a manner inconsistent with the teachings of those references. 8
ITRI agrees with PacBio’s contention that the ’375 patent teaches construction of a 9
CPLM and use single molecule sequencing to provide redundant sequence information 10
by comparing the forward and reverse strand sequences of the CPLM. Opp. at 7 (citing 11
Motion at 6, ll. 19-21, 26-29). ITRI admits that the ’375 patent described “consensus 12
sequence determination through the sequencing of both the sense and antisense strand” of 13
a double stranded segment in a circular molecule. Id. (citing Ex. 1011 at 9, ll. 66-67). 14
ITRI contends that, in 2009, the term “consensus sequence” meant a sequence determined 15
from agreement between related positions in a set of sequence data and that confirming 16
sequence data from one strand using its consistency (i.e., its proper Crick-Watson 17
matching) was conventional in the field. Opp. at 8 (citing Ex. 2048, ¶¶ 48, 54; Ex. 2013, 18
1:21-27 (“Forward and reverse strand sequencing provides the researcher with more 19
information and allows the researcher to evaluate the quality and reliability of the data 20
from both strands. If the bases on both strands complement each other as expected, then 21
this helps to confirm the reliability of the sequence information”). Consequently, argues 22
ITRI, one of ordinary skill would have known that obtaining a consensus sequence with 23
forward and reverse strand sequence data involves looking for complementarity, i.e., 24
Interference 105,970

31

agreement or redundancy, and discarding disagreements. Id. (citing Ex. 2048, ¶¶ 48-49, 1
54). 2
Therefore, argues ITRI, the ’375 patent’s teachings to increase sequencing 3
accuracy by disregarding disagreements between the two strands would have led one 4
skilled in the art away from using disagreements between forward and reverse strands as 5
set forth in the methods of ITRI’s claims 1, 24, and 26. Opp. at 9. (citing Ex. 2048; 6
¶¶ 55, 57, 63). According to ITRI, a method that disregards any mismatch, such as the 7
one disclosed in the ’375 patent, is the antithesis of the method recited in claims 1, 24, 8
and 26 of ITRI’s ’630 patent. Id. 9
ITRI additionally argues PBC failed to explain why Laird or Matsumura would 10
have countered that teaching to arrive at the claimed methods, and that one of ordinary 11
skill in the art would have had no reason to modify the teaching of the ’375 patent 12
to reach the methods of any of claims 1, 24, or 26. ITRI contends that neither Laird nor 13
Matsumura teaches or suggests determining the position of a modified base by comparing 14
sequences of forward and reverse strands or by using a disagreement between those 15
strands. Opp. at 10 (citing Ex. 2048, ¶¶ 59-62). 16
With respect to claim 23, ITRI argues that PacBio’s argument fails because PacBio 17
has failed to allege that step (h) was disclosed in any of the references. Opp. at 11. ITRI 18
argues that step (h) recites, in part, “accepting or rejecting at least four of the repeats of 19
the sequence of the nucleic acid sample contained in the sequence data according to the 20
scores of one or both of the sequences of the inserts immediately upstream and 21
downstream of the sample sequences ….” Id. at 11-12. According to ITRI, PacBio never 22
mentions accepting or rejecting repeats of a sample according to the score of an insert in 23
its argument about claim 23. Id. at 12 (citing Motion at 12-13, ll. 24-10). Instead, argues 24
Interference 105,970

32

ITRI, PacBio refers to “scoring methods whereby calls of iterative reads or from 1
complementary sequences are scored by the number of occurrences.” Id. (citing Motion 2
at 13, ll. 8-9). ITRI argues that the passages PacBio cites in Exs. 1011 and 1006 also do 3
not disclose accepting or rejecting repeats of a sample according to the score of an insert. 4
Id. 5
With respect to claims 27 and 28, ITRI argues that the ’375 patent neither teaches 6
nor suggests using a nucleotide analog that discriminates between a base and its modified 7
form or determining the positions of modified bases. Opp. at 13 (citing Ex. 2048, ¶ 66). 8
According to ITRI, PacBio provides only an allegedly conclusory allegation that a person 9
“of ordinary skill would have been motivated to do so to increase the accuracy of the 10
sequence determination, given the importance of DNA methylation in the regulation of 11
gene expression, genomic imprinting and X-chromosome inactivation.” Id. (citing 12
Motion at 14, ll. 11-14). ITRI argues that the ’375 patent provide increased accuracy, 13
and PacBio has not shown or argued that its alleged combination would provide any 14
further increase. Id. (citing Motion at 14, ll. 7-18; Ex. 2048, ¶ 56). Furthermore, argues 15
ITRI, even if there were a marginal increase in accuracy under some circumstances, 16
PacBio has not shown that this motivation would outweigh the expected problem of 17
polymerase enzyme inhibition. Id. (citing Motion at 13-14, ll. 15-18). 18
ITRI argues that PacBio’s alleged motivation requires that one of ordinary skill 19
must have understood that combining the ’614 patent with the ’375 patent would have 20
“increase[d] the accuracy of the sequence determination” beyond what the ’375 patent 21
already provided. Opp. at 14. However, argues ITRI, PacBio’s statement alleging this 22
motivation lacks support from any evidence or even further argument. Id. 23
Interference 105,970

33

We are persuaded by PacBio’s arguments and conclude that the claims of the 1
‘630 patent are obvious over the combination of the ’375 patent, Laird and Matsumara. 2
Both parties agree that the ’375 patent teaches construction of a CPLM and the use of 3
single molecule sequencing to provide redundant sequence information by comparing the 4
forward and reverse strand sequences of the CPLM. See Opp. at 7; Motion at 6. 5
Furthermore both parties agree that the ’375 patent described “consensus sequence 6
determination through the sequencing of both the sense and antisense strand” of a double 7
stranded segment in a circular molecule. Id. Consequently, this leaves remaining only 8
limitation c of claim 24 of the ’630 patent, which ITRI has chosen as an exemplar. 9
Limitation c of claim 24 recites: 10
c. determining the sequence of the double stranded nucleic acid sample and 11
the position of the at least one modified base in the sequence of the 12
double stranded nucleic acid sample by comparing the sequences of the 13
forward and reverse strands of the circular pair-locked molecule wherein 14
at least one modified base in the double-stranded nucleic sample is paired 15
with a base having a base pairing specificity different from its preferred 16
partner base. 17
18
Ex. 1002. ITRI’s argument that one of ordinary skill would have known that obtaining a 19
consensus sequence with forward and reverse strand sequence data involves looking for 20
complementarity, i.e., agreement or redundancy, and discarding disagreements is 21
countered by the teachings of Laird. 22
Laird teaches that methylated and unmethylated CPG dyads in a bisulfite-treated 23
DNA sequence can be identified by the matching or mismatching of cytosines in the 24
forward and reverse strand sequence data. Laird recites: 25
Bisulfite conversion was used because it provides information on the 26
methylation state of individual cytosines by converting cytosine (but not 5-27
Interference 105,970

34

methylcytosine) to uracil, and subsequently to thymine upon PCR 1
amplification (28, 29). Bisulfite conversion also reduces base pair 2
complementarity within the hairpin, a reduction that we found useful for 3
PCR amplification, as well as for high-fidelity cloning and sequencing of 4
hairpins containing unmethylated cytosines. 5
6
Ex. 1012 at 2. Moreover, the purpose of Laird’s study is to examine the stability of 7
cytosine methylation in CpG/CpG
32
dyads, which it achieves by comparing the 8
forward and reverse strands of the DNA sample and identifying “mismatched” 9
base pair configurations. Figure 2D of Laird is illustrative: 10
11
Figure 2D depicts forward and reverse strands of bisulfite-PCR treated DNA 12
13
The highlighted sequences of the paired forward and reverse strands depicted in 14
Figure 2D of Laird depict CpG sequences that have either been transformed to blue-15
highlighted thymine-guanine pairs (indicating non-methylated cytosine was originally at 16
that locus) and red-highlighted cytosine-guanine pairs (indicating that the unconverted 17
cytosine is 5-methylcytosine at that locus). Although the principle aim of Laird’s 18
teaching is to investigate the stability of methylated CpG/CpG dyads, the manner in 19
which this aim is achieved is by comparing the sequences of the converted forward and 20

32
CpG refers to covalently linked (i.e. positioned adjacently in a strand) cytosine and
guanine bases.
Interference 105,970

35

reverse DNA strands and identifying mismatched (i.e., thymine-guanine) base pairs in the 1
sequence. See Ex. 1012 at 3 (“In sequences with efficient conversion of non-CpG 2
cytosines (97% conversion), the four informative hemimethylated CpG/CpG dyads (i.e., 3
those between the primer sites) showed no conversion of the methylated CpGs (0/24), 4
and 100% conversion (24/24) of the four complementary, unmethylated CpGs”). 5
ITRI argues that the ’375 patent would teach a person of ordinary skill in the art to 6
look for agreement in the sequencing of base pairs in the forward and reverse strands and 7
to discard disagreements (i.e., mismatched base pairs). ITRI implicitly argues, therefore, 8
that the ’375 patent teaches away from the methods taught by the ’630 patent. 9
We disagree. A reference teaches away when “a person of ordinary skill, upon 10
reading the reference, would be discouraged from following the path set out in the 11
reference, or would be led in a direction divergent from the path that was taken by the 12
applicant.” In re Gurley, 27 F.3d 551, 553 (Fed. Cir. 1994). ITRI points to no teaching 13
in the ’375 patent, or the other cited prior art references, that would explicitly or 14
implicitly discourage a person of ordinary skill from looking for base pair disagreements 15
as well as base pair agreements and, as we have related supra, Laird explicitly teaches 16
looking for guanine-thymine mismatches as evidence of non-methylated cytosine in a 17
forward or reverse strand locus. Moreover, observing disagreement between base pairs is 18
inherent in observing base pair agreement, i.e., base pair agreement/disagreement under 19
the Crick-Watson rules is a binary choice; the bases either or properly matched or they 20
are not. Thus, when observing whether the base pairs agree, one necessarily observes 21
whether they disagree. See In re Cruciferous Sprout Litig., 301 F.3d 1343, 1349 (Fed. 22
Cir. 2002) (Inherency can only be established when “prior art necessarily functions in 23
accordance with, or includes, the claimed limitations”). 24
Interference 105,970

36

With respect to claim 23, we are persuaded that the cited prior art references teach 1
or suggest limitation h, as ITRI maintains. Limitation h recites: 2
h. accepting or rejecting at least four of the repeats of the sequence of the 3
nucleic acid sample contained in the sequence data according to the 4
scores of one or both of the sequences of the inserts immediately 5
upstream and downstream of the sample sequences, subject to the 6
condition that at least one sample sequence in each orientation is 7
accepted. 8
9
Ex. 1002. The ’375 patent recites: 10
Although referred to herein as comparing or assembling the sequence data 11
from multiple reads of a given sequence, and/or from the sense and antisense 12
strands of the sequence, it will be appreciated that any method of assigning a 13
consensus determination to a particular base call from multiple reads of that 14
position of sequence, and/or to a provide an overall consensus sequence for 15
that segment, will be envisioned and encompassed by the term “compare.” 16
17
Ex. 1011, col. 11, ll. 40-47. The ’375 patent thus explicitly teaches determining 18
consensus from “multiple reads of that position of sequence.” Moreover, Laird teaches 19
determining sequences and agreement/disagreement of base pairs at given loci. See Zon 20
Decl. 2 at Fig. 2. Although neither reference explicitly requires that at “least four of the 21
repeats” are accepted or rejected, the very nature of the teachings of the ’375 patent and 22
Laird will produce at least four repeats. The ’630 patent defines “repeat” as: 23
A repeat or repeat sequence is a sequence that occurs more than once in a 24
nucleic acid. When repeats are present in a nucleic acid molecule, all 25
instances of the sequence, including the first instance, are considered 26
repeats. Repeats include sequences that are reverse complements of each 27
other, such as occur in a circular pair-locked molecule. 28
29
Ex. 1002, col. 10, ll. 40-44. Thus a single iteration of the formation of the CPLM 30
and single replication of the double stranded DNA sample will yield four repeats: 31
Interference 105,970

37

the forward and reverse strands of the original sample and the complementary 1
strands formed by polymerization. See, e.g., Ex. 1011, col. 20, ll. 50-62, Figs. 5, 6. 2
We therefore conclude that the combined cited prior art references teach 3
limitation h of claim 23. 4
With respect to claims 27 and 28, we are similarly not persuaded by ITRI’s 5
argument that the prior art fails to teach or suggest the limitation reciting “[a]t least 6
one nucleotide analog that discriminates between a base and its modified form is 7
used to obtain sequence data comprising at least one position wherein the at least 8
one differentially labeled nucleotide analog was incorporated.” ITRI argues that 9
the ’375 patent provides increased accuracy, and asserts that PacBio has not shown 10
or argued that its alleged combination would provide any further increase. Opp. at 11
13. 12
Laird teaches such a method or discriminating between methylated and non-13
methylated cytosines, and comments extensively on the developmental importance 14
of cytosine methylation, particularly with respect to epigenetic cellular inheritance. 15
See, e.g., Ex 1025 at 1. We find that a person of ordinary skill in the art would 16
know that the teachings of the ’375 patent could be combined with the teachings of 17
Laird to increase the accuracy of the method by discriminating between methylated 18
and non-methylated cytosines, and we conclude that the combined cited prior art 19
references tech or suggest all of the limitations of claims 27 and 28. 20
“The combination of familiar elements according to known methods is likely 21
to be obvious when it does no more than yield predictable results.” KSR Int’l Co. 22
v. Teleflex Inc., 550 U.S. 398, 416 (2007). The combined cited prior art references 23
teach or suggest all of the limitations of the claims of the ’630 patent and we 24
Interference 105,970

38

conclude that a person of ordinary skill in the art would be motivated to combine 1
the references to increase the accuracy of the invention, particularly with respect to 2
discriminating between methylated and demethylated cytosine bases. PacBio’s 3
Motion 3 for judgment of unpatentability under 35 U.S.C. § 103(a) against ITRI’s 4
’630 patent is consequently GRANTED. 5
6
7
IV. ITRI’S Motion 2 to deny PacBio the benefit of its Application 8
No. 61/201,551
33
9
10
ITRI’s Motion 2 seeks judgment that Pacific Biosciences is not entitled to 11
the benefit of its Application No. 61/201,551 (the “’551 application”)
34
(Ex. 2026) 12
which was filed December 11, 2008. Motion
35
at 1. PacBio’s involved 13
US Applications 13/633,673 and 13/930,178 claim the benefit of the 14
’551 application. Redecl.
36
at 3. According to ITRI, the ’551 application fails to 15
provide any disclosure indicating that PacBio invented an embodiment within the 16
scope of the Count. Id. 17
Specifically, ITRI alleges that, in contrast to the method disclosed in ITRI’s 18
’630 patent, claim 24 of which constitutes the sole count of this interference, the 19
’551 application does not use a mismatch or a disagreement between bases pairs in the 20
forward and reverse strands. Motion at 5 (citing Levy First Decl., ¶ 36) (“Levy Decl. 1”). 21

33
Per the March 31, 2014 order of the Board, ITRI’s Motion 3 is deferred until the
priority phase of the interference. See Paper No. 106.
34
Paper No. 84
35
Paper No. 59
36
Paper No. 28
Interference 105,970

39

Rather, contends ITRI, the ’551 application describes an approach using a distinct signal 1
that occurs when a nucleotide is incorporated across from a modified base during 2
sequencing-by-synthesis. Id. (citing Ex. 2026, ¶¶ 16-17; Ex. 2003, ¶ 33). 3
ITRI argues that, when interpreted in light of ITRI’s ’630 Specification, the 4
Count requires using a disagreement between forward and reverse strand 5
sequences of the CPLM to determine the position of a modified base. Motion at 6 6
(citing Levy Decl. 1, ¶ 30). The count of the instant interference recites, in 7
relevant part: 8
c. determining the sequence of the double stranded nucleic acid sample and 9
the position of the at least one modified base in the sequence of the 10
double stranded nucleic acid sample by comparing the sequences of the 11
forward and reverse strands of the circular pair-locked molecule wherein 12
at least one modified base in the double-stranded nucleic sample is paired 13
with a base having a base pairing specificity different from its preferred 14
partner base [i.e., is “mismatched”]. 15
16
Ex. 1002. Therefore, argues ITRI, the count comprises sequencing the forward and 17
reverse strands of a CPLM containing a mismatched modified base, such as a uracil 18
paired to guanine, and that obtaining these sequences reveals a disagreement—the uracil 19
directs incorporation of an adenine into the growing strand, whereas the guanine directs 20
incorporation of a cytosine into the opposite growing strand. Motion at 7 (citing Levy 21
Decl. 1, ¶¶ 28-29). In the limitation of the count recited supra, the disagreement is used 22
to conclude that a modified base was present at the position of the disagreement 23
(specifically, that the modified base was a cytosine that has been deaminated to uracil). 24
Id. (citing Levy Decl. 1, ¶ 30). 25
ITRI argues that the ’551 application does not describe an embodiment of this 26
limitation of the count. Motion at 8. Specifically, argues ITRI, the ’551 application lacks 27
Interference 105,970

40

a written description of “determining … the position of the at least one modified base … 1
by comparing the sequences of the forward and reverse strands of the circular pair-locked 2
molecule, wherein at least one modified base in the double stranded nucleic acid sample 3
is” mismatched. Id. (citing Levy Decl. 1., ¶¶ 37, 39). ITRI contends that the 4
’551 Specification lacks express description of comparing forward and reverse strand 5
sequences that disagree at a modified base position due to a mismatch involving the 6
modified base, and using that mismatch to determine the modified base position. Id. at 9 7
(citing Levy Decl. 1, ¶¶ 36-37). 8
According to ITRI, the only mention of using the reverse strand sequence in a 9
modified base determination in the ’551 application recites “For example, sequence reads 10
from the sense or ‘forward’ strand can be compared to sequence reads from the antisense 11
or ‘reverse’ strand for the same nucleic acid template to further validate the existence of 12
one or more modified bases in the template nucleic acid.” Id. at 9 (quoting Ex. 2026, ¶ 13
17; citing Levy Decl. 1, ¶ 36). ITRI contends that neither that sentence, nor its context, 14
conveys using a disagreement between the forward and reverse strand sequences to 15
determine the modified base position and, therefore, the ’551 specification does not 16
expressly describe using such a disagreement to determine the position of the modified 17
base. Id. (citing Levy Decl. 1, ¶ 36). 18
Similarly, argues ITRI, neither paragraph 18 nor 21 of the ’551 application 19
supports the language of the count reciting a mismatch in the sample. Motion at 9. 20
(citing Ex. 2010 at 5-6). According to ITRI, Paragraph 18 refers to several modifications 21
“to improve the signal discrimination approaches provided herein,” including the use of 22
modified nucleotides to create the new strand during sequencing “to enhance the 23
discrimination of various bases” but does not disclose methods of discriminating 24
Interference 105,970

41

modified bases that involve a mismatch in the sample. Id. (citing Ex. 2026, ¶ 18). ITRI 1
observes that paragraph 21 recites: “[a]ltered nitrogenous bases may exhibit different 2
base pairing and/or base stacking properties as compared to native bases. For example, 3
‘universal bases’ can pair with all four native bases.” Id. at 10 (quoting Ex. 2026, ¶ 21). 4
ITRI contends that different base pairing properties, particularly those of universal bases, 5
is not the same thing as a mismatch. Id. (citing Levy Decl. 1, ¶ 70). 6
Further, ITRI notes that PacBio’s Application No. 61/099,696
37
(the 7
“’696 application,” filed September 24, 2008) (Ex. 2015), which is incorporated by 8
reference into the ’551 application, recites “comparing sequence data from each strand to 9
determine a consensus sequence for the double stranded segment” where the double 10
stranded segment was sequenced as part of a circular molecule. Motion at 11 (quoting 11
Ex. 2015, p. 30). ITRI also observes that the ’696 application recites “as with circular 12
templates, the template configurations of the invention provide single molecular 13
consensus sequences, where sequencing a given template provides duplicative data of the 14
sequence information obtained, and thereby improves accuracy over linear templates by 15
providing confirming or consensus sequence data on a given sequence.” Id. (quoting 16
Ex. 2015, ¶ 33). However, argues ITRI, the consensus sequence is obtained by obtaining 17
duplicative or redundant data that shows agreement between the forward and reverse 18
strands, and any mismatches at a given position in the sequencing data will be overruled 19
by the more prevalent redundant agreement shown in the data. Id. (citing Levy Decl. 1, 20
¶¶ 47-48). 21
ITRI also argues that the ’551 application’s use of the term “validation” 22
demonstrates that the ’551 inventors did not contemplate that the forward and reverse 23

37
Paper No. 75
Interference 105,970

42

strand sequences would show a disagreement. Motion at 12 (citing Ex. 2026). 1
“Validation” of a modified base with reverse strand sequence data, contends ITRI, 2
implies that the data show complementarity, i.e., agreement or redundancy, with the 3
modified base. Id. (citing Levy Decl. 1, ¶ 61). ITRI argues that as of 4
December 11, 2008, sequencing approaches for identifying modified bases using a 5
sequence comparison would compare one strand of a sample sequence to a reference 6
sequence, not the reverse strand of the nucleic acid. Id. (citing Ex. 2020, p. 206; 7
Ex. 2003, ¶ 60). 8
Finally, ITRI argues that whereas the count requires a mismatched modified base 9
in the sample and determining the position of the modified base using a disagreement 10
between forward and reverse sequences, statements about its inventive detection of 11
modified bases, particularly uracil, in the ’551 application explicitly exclude methods of 12
determining the position of a modified base using a disagreement between the forward 13
and reverse strand sequences at the modified base position. Motion at 14 (citing Levy 14
Decl. 1, ¶ 74). 15
ITRI points to the ’551 applications recitation that uracil “can be found in DNA … 16
as the result of bisulfite-conversion of cytosine in a common protocol used to 17
discriminate methylated cytosine through DNA sequencing. Unlike this common 18
protocol, the methods herein directly detect the modified base rather than relying on the 19
similarity of uracil to thymine.” Id. at 14-15 (quoting Ex. 2026, ¶ 23). ITRI contends 20
that the methods of the ’551 application rely on detecting some behavior of uracil that 21
differs from that of thymine, such as a change in the kinetics of sequencing-by-synthesis 22
data, a difference in behavior of a polymerase when it encounters uracil, or statistical 23
Interference 105,970

43

differences in error rate frequencies, none of which uses a mismatch to detect uracil. Id. 1
at 15 (citing Levy Decl. 1, ¶¶ 73-74, Ex. 2026, ¶¶ 16-17, 28, 29). 2
PacBio responds that ITR’s arguments are based on an overly restrictive and 3
incorrect interpretation of the count and that a step of “using a disagreement” is not 4
recited in the count. Opp. at 4. According to PacBio, nowhere in the count is it recited 5
that the method requires using a disagreement to determine the position of a modified 6
base, let alone using the disagreement at all, or does the count recite that “the at least one 7
modified base” detected by comparing the forward and reverse strands is the same 8
modified base as that recited in the last limitation of the count. Id. at 5. 9
PacBio contends that, upon considering the ’630 Specification in its entirety, one 10
of ordinary skill in the art would have understood that the Specification teaches detecting 11
the position of a modified base via the method of the count regardless of whether the base 12
is mismatched. Opp. at 5 (citing Ex. 1002, cols. 34-35, ll. 60-20; Zon Decl. 1
38
, ¶¶ 43-49, 13
52-53, 57, 61 (Ex. 1024); Zon Decl. 2, ¶¶ 23-27). PacBio points out that in the context of 14
a bisulfite treated DNA template, it is the non-mismatched 5-methylcytosine positions 15
that one of ordinary skill in the art generally seeks to detect and that the position of a 5-16
methylcytosine can be detected in a bisulfite-treated CPLM by comparing the forward 17
and reverse strands of a CPLM and detecting a normal Crick-Watson C-G base pair. Id. 18
(citing Ex. 1002, col. 2, ll.14-23, cols. 34-35; ll. 60-20; Zon Decl. 1, ¶ 47; Zon Decl. 2, 19
¶¶ 18, 24-26). PacBio argues that ITRI’s proposed construction of the count would 20
exclude the detection of a modified base that is thus not mismatched. Id. 21
PacBio also disputes ITRI’s interpretation of the claim language to mean that 22
determining the position of a modified base does not encompass validation of the position 23

38
Paper No. 136
Interference 105,970

44

of the modified base. Opp. at 6. ITRI contends that, although “determining” is not 1
defined in the ’630 patent, its use in the specification is 27 consistent with the ordinary 2
and customary meaning of the term. Id. PacBio points to Example 6 of the 3
’630 Specification, which describes comparing the forward and reverse strands of a 4
CPLM to “determine the sequence of the nucleic acid sample and the position of the at 5
least one 5-methylcytosine” and that the template nucleic acid can also include a 6
mismatched base, for example a uracil mismatched with a guanine, due to bisulfite. Id. at 7
6-7 (quoting Ex. 1002, col. 34-35 ll. 60-20). PacBio contends that this comparing step is 8
a validation of the initial determination of the position of a 5-methylcytosine residue, 9
because the position of the 5-methylcytosine residue is immediately determined upon 10
obtaining a G read in the sequencing data in the absence of a comparing step. Id. at 7 11
(citing Zon Decl. 2, ¶¶ 24, 26). Accordingly, argues PacBio, the act of comparing the 12
two strands to determine the position of a C-G matched pair is a validation of the 13
determination of the position of the 5-methylcytosine, since the position of the 5-14
methylcytosine is initially determined from reading either the forward or reverse strand 15
alone. Id. (citing Zon Decl. 1, ¶ 47; Zon Decl. 2, ¶ 26). 16
PacBio next argues that CPLM templates and bisulfite treatment are described in 17
the ’551 application in the same paragraph that also teaches comparing the forward and 18
reverse strands of the CPLM to determine the presence of a modified base. Opp. at 10. 19
According to PacBio, one of ordinary skill in the art would have readily understood that 20
the sequencing methods taught in the ’551 application could be carried out on the 21
bisulfite treated template also described therein and, further, would have also understood 22
that bisulfite treatment could be applied to CPLM templates to create mismatched 23
modified bases (i.e., uracil paired with guanine) and would also leave the base pairing 24
Interference 105,970

45

specificity of 5-methylcytosine (a modified base) unchanged. Id. (citing Zon Decl. 1, 1
¶¶ 40-42). 2
PacBio also contends that the ’551 application further states that a comparison of 3
data reads of the forward and reverse strands “can provide additional information for 4
discriminating modified from unmodified 6 nucleotides.” Opp. at 11 (citing Ex. 1004, 5
¶ [0017]). PacBio contends that one of ordinary skill in the art would have understood 6
that the comparison step taught at ¶ [0017] of the ’551 application could be used to 7
determine the position of a mismatched modified base, i.e., uracil, in a bisulfite treated 8
CPLM, initially determined by a separate mechanism such as incorporation signal 9
strength, signal duration or signal frequency. Id. (citing Zon Decl. 1 ¶¶ 47, 61). 10
Accordingly, argues PacBio, when ¶ [0017] of the ’551 application is read and 11
interpreted in the context of the ’551 application as a whole, one of ordinary skill in the 12
art would understand that it describes an embodiment where comparing the forward and 13
reverse strands of a bisulfite-treated CPLM is used to determine the position of the 14
modified base (e.g., uracil or 5-methylcytosine) determined initially by the other readout 15
mechanisms described in 16 the ’551 application, such as signal duration and signal 16
strength. Opp. at 11 (citing 1004, ¶¶ [0016]-[0017], [0026]; Zon Decl. 1, ¶¶ 47, 49, 61). 17
We are not persuaded by ITRI’s arguments. As we related supra, a person of 18
ordinary skill in the art as of 2009 would have possessed a Ph.D. in molecular biology, or 19
a related field, and 1-2 years’ relevant postdoctoral experience and would, therefore, 20
possess a sophisticated understanding of the molecular biology of nucleic acids and their 21
synthesis. 22
The ’551 application explicitly defines “modified bases” as including, inter alia, 23
“methylated bases; bisulfite-converted bases” that “may exhibit different base pairing 24
Interference 105,970

46

and/or base stacking properties as compared to native bases.” Ex. 2026, ¶ [0021]. 1
Moreover, the ’551 application teaches that “uracil can be found in DNA as the result of 2
… of bisulfite-conversion of cytosine in a common protocol used to discriminate 3
methylated cytosine through DNA sequencing.” Id., ¶ [0023]. We consequently find that 4
one of ordinary skill in the art would have known that treating DNA with bisulfite, 5
according to a protocol well-known in the art, would result in the conversion of cytosine 6
(but not 5-methylcytosine) to uracil, which pairs, according to the Crick-Watson rules, 7
preferentially with adenine rather than guanine. See, e.g., Levy Decl. 1, ¶ 21. 8
The ’551 application also teaches that “redundant sequence information can be 9
generated by sequencing the same template nucleic acid multiple times by, e.g., 10
sequencing the same template nucleic acid molecule repeatedly or by sequencing 11
multiple copies of the same template nucleic acid” and that: 12
Such redundant sequence information can provide additional information for 13
discriminating modified from unmodified nucleotides. For example, 14
sequence reads from the sense or “forward” strand can be compared to 15
sequence reads from the antisense or “reverse” strand for the same nucleic 16
acid template to further validate the existence of one or more modified bases 17
in the template nucleic acid. In some preferred applications, the forward and 18
reverse strands are sequenced in a single template nucleic acid. 19
20
Ex. 2026, ¶ [0017]. The ’551 application also teaches: 21
Sufficiently redundant sequence data would facilitate detection of such a 22
change in misincorporation rate. In particular, uracil (U) has a propensity to 23
base pair with adenine (A) and somewhat with guanine (G), so with 24
redundant sequencing data and knowledge of the misincorporation 25
frequencies, the presence of U can be detected in a template nucleic acid. 26
For example, both A and G would be incorporated at an “unmethylated” C 27
site (bisulfite-converted to U) while only G would be incorporated at a C site 28
that was not bisulfite-converted to U 29
Interference 105,970

47

1
Id. We consequently find that the ’551 application teaches the portion of limitation c of 2
the count that recites: “determining the sequence of the double stranded nucleic acid 3
sample and the position of the at least one modified base in the sequence of the double 4
stranded nucleic acid sample by comparing the sequences of the forward and reverse 5
strands of the circular pair-locked molecule.” The ’551 application teaches “determining 6
the sequence of the double stranded nucleic acid sample” by teaching “sequence reads 7
from the … ‘forward’ strand can be compared to sequence reads from the … ‘reverse 8
strand for the same nucleic acid template.” 9
ITRI’s declarant, Dr. Levy, opines that “the ’551 application discloses different 10
ways of determining a modified base position that do not use a mismatch between two 11
strands.” Levy Decl. 1, ¶ 33. According to Dr. Levy, “determining a modified base 12
position in methods of the ’551 application must rely on a distinct signal from the 13
sequencing of the modified base itself, that does not occur with an unmodified base one 14
wishes to distinguish it from.” Id. Dr. Levy contends that the passages from paragraph 15
17 of the ’551 application cited supra “does not describe using a mismatch in a sample or 16
CPLM involving a modified base to identify a modified base’s position.” Id., ¶ 36. 17
Moreover, Dr. Levy opines that the ’551 application also does not inherently 18
describe determining the position of a modified base by using a mismatch involving a 19
modified base in a sample or CPLM. Id., ¶ 39. According to Dr. Levy, if there were a 20
mismatched modified base, the information from the forward and reverse sequences 21
would not be redundant, because it would allow one to determine that the strands 22
disagree—something that would not be possible from sequencing either strand alone. Id., 23
¶ 42. Dr. Levy states further that: 24
Interference 105,970

48

One of ordinary skill would not understand the sentence about comparing 1
forward and reverse strands to necessarily involve a non-redundant 2
disagreement (revealing a mismatch) because it begins with the words “For 3
example” and follows three sentences in a row that each discuss “redundant 4
sequence information.” Ex. 2026, ¶ 17. Thus, the ’551 application indicates 5
that comparing the sequences of the forward and reverse strands is a way to 6
use the redundancy discussed in the previous sentences. 7
8
Id., ¶43. Dr. Levy concludes: 9
[T]he mention of comparing strands in the ’551 application would not be 10
understood as requiring a mismatch or disclosing the use of a mismatch to 11
identify a modified base position. Instead, a consensus sequence could be 12
generated for a molecule that contains a modified base such as 5-13
methylcytosine. One could do so by making and sequencing a circular 14
molecule, and then comparing forward and reverse strand sequences 15
according to the example in the ’696 application. A match, i.e., redundant 16
sequence information, involving the modified base (e.g., resulting from a 17
pairing of the 5-methylcytosine to guanine) would aid in constructing the 18
consensus sequence, because the corresponding forward and reverse strand 19
sequence data agree. 20
21
Id., ¶ 52. 22
We are not convinced. The ’551 application teaches that “sequence reads from the 23
… “forward” strand can be compared to sequence reads from the … “reverse” strand for 24
the same nucleic acid template to further validate the existence of one or more modified 25
bases in the template nucleic acid. Ex. 2026, ¶ [0017]. The ’551 application thus teaches 26
comparing the sequences of forward and reverse strands. ITRI argues that this teaches 27
for detecting matched base pairs, but not mismatched base pairs; Dr. Levy specifically 28
states that a match resulting from the pairing of unconverted 5-methylcytosine with 29
guanine, would aid in constructing the consensus sequence, because the forward and 30
reverse strand data would agree. Levy Decl. 1, ¶ 52. 31
Interference 105,970

49

However, given the relatively high level of skill of the average artisan, as we have 1
defined it, it is not reasonable to believe that such a person would not recognize a 2
mismatch occurring between a modified base, i.e., cytosine converted to uracil by 3
bisulfite treatment (as taught by the ’551 application) paired with guanine, and not 4
understand from that mismatch that “at least one modified base in the double-stranded 5
nucleic sample is paired with a base having a base pairing specificity different from its 6
preferred partner base” as required by the disputed limitation. To reason otherwise, to 7
contend that a person of ordinary skill would only look at “correct” base pair matches and 8
discard, without thought, the mismatched pairs, is not reasonable because the person of 9
ordinary skill would know, a priori, that bisulfite-converted uracil would be expected to 10
be paired, incorrectly, with the guanine that would normally be paired with the pre-11
converted cytosine. In other words, a person of ordinary skill would expect to see 12
mismatched uracil-guanine base pairs in a DNA sample strand that had been modified by 13
bisulfite-conversion. Ex. 2026, ¶ [0017] 14
The ’551 application teaches, by way of example, “sequence reads from the sense 15
or “forward” strand can be compared to sequence reads from the antisense or “reverse” 16
strand for the same nucleic acid template to further validate the existence of one or more 17
modified bases in the template nucleic acid.” Ex. 2026, ¶ [0017]. We do not agree with 18
ITRI that the use of “validate” in this passage of the ’551 application necessarily requires 19
that the analysis of base pair only validate that the base pairs agree with the Crick-20
Watson rules. To the contrary, we find that finding a mismatched uracil (or thymine)-21
guanine pair would “validate”, commonly understood to mean “support or corroborate,” 22
the existence of a modified base (i.e., a bisulfite-modified cytosine) in the template 23
Interference 105,970

50

nucleic acid on a sound or authoritative basis (viz., knowledge of the bisulfite conversion 1
reaction). 2
ITRI has not shown that the ’551 application, when viewed in light of the ’630 3
Specification, fails to disclose “comparing forward and reverse strands” “to further 4
validate the existence of one or more modified bases in the template nucleic acid.” We 5
consequently determine that ITRI has not met its burden of showing why we should deny 6
PacBio the benefit accorded its’551 application. ITRI’s Motion 2 is DENIED. 7
8
V. Conclusion 9
For the reasons set forth above, PacBio’s Motion 1 to rescind the benefit accorded 10
to ITRI for its’3l3 application for failure to provide a constructive reduction to practice of 11
Count 1 is DENIED. PacBio’s Motion 2 seeking judgment of unpatentability under 12
35 U.S.C. § 103(a) against ITRI’s ’630 patent is GRANTED. ITRI’S Motion 2 to deny 13
PacBio the benefit of its ’551 application is DENIED. 14
Interference 105,970

51

Attorney for Pan: 1
2
David S. Forman, Ph.D. 3
M. Paul Barker 4
Adam M. Breier, Ph.D. 5
Finnegan, Henderson, Farabow, Garrett & Dunner, LLP 6
David.forman@finnegan.com 7
Paul.barker@finnegan.com 8
Adam.breier@finnegan.com 9
10
Attorney for Flusberg: 11
12
Michael S. Tuscan, Ph.D. 13
Bonnie Weiss McLeod, Ph.D. 14
Cooley, LLP 15
mtuscan@cooley.com 16
bweissmcleod@cooley.com 17
zpatdocketing@cooley.com 18
19
20