Sunil Kumar et al.

Patent Trials and Appeals Board

Sep 5, 2019

14363436 - (D) (P.T.A.B. Sep. 5, 2019)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
14/363,436 06/06/2014 Sunil Kumar 2011P01782WOUS 5322
24737 7590 09/05/2019
PHILIPS INTELLECTUAL PROPERTY & STANDARDS
465 Columbus Avenue
Suite 340
Valhalla, NY 10595
EXAMINER
BRUSCA, JOHN S
ART UNIT PAPER NUMBER
1631
NOTIFICATION DATE DELIVERY MODE
09/05/2019 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
katelyn.mulroy@philips.com
marianne.fox@philips.com
patti.demichele@Philips.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte SUNIL KUMAR, RANDEEP SINGH,
BISWAROOP CHAKRABARTI and SUBODH KUMAR1
____________

Appeal 2019-002987
Application 14/363,436
Technology Center 1600
____________

Before ULRIKE W. JENKS, TIMOTHY G. MAJORS, and
MICHAEL A. VALEK, Administrative Patent Judges.

VALEK, Administrative Patent Judge.

DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134(a) involving claims to a
method and apparatus for identifying and calling variants in an assembled
genetic sequence, which have been rejected as directed to patent-ineligible
subject matter. We have jurisdiction under 35 U.S.C. § 6(b).
We REVERSE.

1 Appellants identify the real party in interest as Koninklijke Philips N.V.
App. Br. 1. Herein we refer to the Office Action mailed July 5, 2018 (“Final
Act.”), Appeal Brief filed November 12, 2018 (“App. Br.”), and Examiner’s
Answer mailed December 26, 2018 (“Ans.”).
Appeal 2019-002987
Application 14/363,436

2
STATEMENT OF THE CASE
Appellants’ invention relates to genetic sequencing and, in particular,
an “operation in such an analysis” referred to as “variant identification or
‘calling’.” Spec. 2, l. 3. The Specification explains variant calling as
follows:
This is the operation in which a detected variant is assessed to
identify whether the variant is actually present in the genome of
the patient, or is instead an artifact of or error in the sequencing
and/or assembly processing. The base quality is commonly
measured by a phred like quality score. In the case of Sanger
sequencing, phred quality scores are calculated from
spectrogram data by calculating parameters for the sequenced
base such as peak shape and resolution, and comparing these
values with an empirically developed look-up table. . . .
Coverage is a metric of the number of reads, and is often
expressed in a multiplier form. . . . In general, higher coverage
corresponds to higher read reliability as the coverage indicates
redundancy.
Id. at 2, ll. 4–18.
While “[v]ariant calling using a combination of base quality score and
coverage is generally effective,” the Specification states “there are some
disadvantages to this approach.” Id at 4, ll. 18–20. According to the
Specification, “base quality score and coverage depend upon the sequencing
platform and the alignment algorithm” and “sequencing errors tend to
propagate into the variant calling” such that “the variant call error rate may
be substantially higher than the base call error rate predicated by the phred
(or other base quality) score.” Id. at 4, ll. 20–30 (citing published study
where “sequencing accuracies of greater than 99.97% for various next
generation sequencing (NGS) platforms translated to variant calling
accuracies of below 99%, and below 95% for some sequencing platforms”).
Appeal 2019-002987
Application 14/363,436

3
Moreover, “increasing the coverage is time consuming and incurs greater
reagent costs, and may not reduce the variant call error rate to acceptable
levels.” Id. at 5, ll. 2–3.
Appellants purport to overcome these disadvantages by incorporating
“assessment of properties of the underlying sequencing reads that contain
the possible variant,” that is, “the actual (i.e., raw) output of the sequencer.”
Id. at 5, ll. 23–26. The Specification provides the following description of
this assessment:
The reads are . . . expected to have properties that fall within a
typical range of values for the type of reads under analysis. If
read properties computed for the reads of the genetic sequence
that include the possible variant indicate that those reads are
highly unstable, or deviate from the typical range of values,
then it is likely that those reads are erroneous (that is, contain
errors). Accordingly, the possible variant contained in those
suspect reads may also be erroneous.
Id. at 5, l. 27–6, l. 1. The Specification refers to this assessment of read
properties as a “second-pass acceptance test” because it may be combined
with “‘first pass’ base quality/coverage filtering” to further reduce errors.
Spec. 12, l. 29 – 13, l. 2. The assessed read properties may “be
thermodynamic properties, structural properties, base compositional
properties, or so forth.” Id. at 10, ll. 17–19; see also id. at 12, ll. 24–29
(describing exemplary read properties such as “nucleotide frequency,” “GC
clamp,” “stacking energy”, “DNA denaturation temperature,” “dimer
formation,” “hairpin loop formation,” and “bending stiffness” amongst
others). “A possible variant is ‘called’ (that is, accepted as an actual variant
for clinical purposes)” only if it is classified as accepted based on
consideration of those read properties. Id. at 8, ll. 18–20. According to the
Appeal 2019-002987
Application 14/363,436

4
Specification, “[a] further benefit of the disclosed approach is that, by
combining a plurality of read properties using a classifier, the variant calling
can be tuned during the training of the classifier.” Id. at 6, ll. 1–3.
Claims 1, 2, 7, 14–16, 20, 22, and 23 are on appeal, and can be found
in the Claims Appendix of the Appeal Brief. Claim 1 is illustrative and
reads as follows:
1. A method comprising:
identifying a possible variant in an assembled genetic
sequence comprising aligned sequencing reads;
computing values of at least one read property for
sequencing reads of the assembled genetic sequence; and
calling the possible variant conditional upon the computed
values of the at least one read property for sequencing reads of
the assembled genetic sequence that include the possible variant
satisfying an acceptance criterion;
wherein the possible variant is classified as accepted or
rejected using a classifier operating on the computed values of
the plurality of read properties for the reads that include the
possible variant; and wherein the possible variant is called only
if the possible variant is classified as accepted, wherein
classifying is performed using support vector machine
classification using F-score;
wherein the identifying, computing, and calling are
performed by an electronic data processing device.
App. Br. 21. As shown above, the claim recites “identifying a possible
variant in an assembled genetic sequence comprising aligned sequencing
reads,” i.e., identifying areas where the aligned sequences differ as possible
variants. And following that identification, “using a classifier operating” on
a “plurality of read properties” to classify the possible variant as accepted or
rejected based on “at least one read property” satisfying an “acceptance
criterion” indicative of the reliability of the sequence data “for the reads that
include the possible variant.” The other independent claims recite similar
Appeal 2019-002987
Application 14/363,436

5
limitations. See id. at 22–23 (claims 14 and 16). Claim 7 additionally limits
the method of claim 1 by requiring that “sequencing reads that do not satisfy
base quality score and coverage criteria are discarded and are not included in
the assembled genetic sequence” from which possible variants are identified.
App. Br. 22. Thus, claim 7 further specifies “first pass base
quality/coverage filtering” prior to the “second-pass acceptance test” based
on read properties that is recited in the independent claims. See Spec. 12, l.
29 – 13, l. 2.
Appellants seek review of Examiner’s rejection of claims 1, 2, 7, 14–
16, 20, 22, and 23 under 35 U.S.C. § 101 as directed to patent-ineligible
subject matter.
ANALYSIS
An invention is patent-eligible if it claims a “new and useful process,
machine, manufacture, or composition of matter.” 35 U.S.C. § 101.
However, the Supreme Court has long interpreted 35 U.S.C. § 101 to include
implicit judicial exceptions: “[l]aws of nature, natural phenomena, and
abstract ideas” are not patentable. E.g., Alice Corp. v. CLS Bank Int’l, 573
U.S. 208, 216 (2014).
In determining whether a claim falls within an excluded category, we
are guided by the Supreme Court’s two-step framework, described in Alice.
Id. at 217–18 (citing Mayo Collaborative Servs. v. Prometheus Labs., Inc.,
566 U.S. 66, 75–77 (2012)). In Alice step one, we ask whether the claims
are directed to an exception to patent eligibility, such as an abstract idea or
law of nature. Alice Corp. Pty. V. CLS Bank Int’l, 134 S. Ct. 2347, 2355
(2014). In Alice step two, we examine the elements of the claims to
determine whether they contain an inventive concept sufficient to transform
Appeal 2019-002987
Application 14/363,436

6
the claimed judicial exception into a patent-eligible application. Mayo, 566
U.S. at 71–72 (quoting Alice, 134 S. Ct. at 2355).
The Office recently published revised guidance on the application of
the Supreme Court’s Alice analysis. USPTO’s January 7, 2019
Memorandum, 2019 Revised Patent Subject Matter Eligibility Guidance, 84
Fed. Reg. 50–57 (“Guidance”). According to the Guidance, we look to
whether the claim recites: (1) a judicial exception, such as a law of nature
(Guidance Step 2A, prong 1); and (2) additional elements that integrate the
judicial exception into a practical application (Guidance Step 2A, prong 2).
Only if the claim recites a judicial exemption and does not integrate that
exception into a practical application, do we then examine whether the claim
adds a specific limitation beyond the judicial exception that is not “well-
understood, routine, conventional” in the field (Guidance Step 2B). See
Guidance at 54–56.
Examiner’s Findings and Conclusions
Regarding Alice step one, Examiner determines that claim 1 is
“directed to a judicial exception of determining the validity of a variant in
genetic sequence reads using classifiers based on properties of the sequence
reads.” Final Act. 4. Examiner equates Appellants’ claims to those
determined to recite patent ineligible subject matter in In re BRCA1- and
BRCA2-Based Hereditary Cancer Test Patent Litigation, 774 F.3d 775 (Fed.
Cir. 2014) (“In re BRCA1”) because, in Examiner’s view, “both sets of
claims . . . are directed to an abstract idea of analysis of genetic sequence
data.” Ans. 3; see also Final Act. 8 (finding that Appellants’ claims are
“similar to the abstract idea of comparing information of a sample or test
Appeal 2019-002987
Application 14/363,436

7
subject to a control or target data” at issue in In re BRCA1).2 At Alice step
two, Examiner determines Appellants’ claims merely recite the “additional
element of use of a generic computer to analyze abstract [genetic sequence]
data” and other “conventional computer process[es].” Final Act. 5. For this
reason, Examiner finds “[t]he additional elements [of Appellants’ claims]
alone or in combination do not comprise an inventive concept.” Id. at 7–8.
Appellants’ Contentions
Appellants challenge Examiner’s rejection at both steps of the Alice
analysis. Appellants argue that only the first step of claim 1, i.e.,
“identifying a possible variant in an assembled genetic sequence comprising
aligned sequencing reads” is analogous to the abstract idea in In re BRCA1
and that Examiner “improperly distill[ed] the numerous limitations”
additionally requiring that a possible variant identified by comparison of the
assembled test sequence to an aligned reference sequence be further
classified as either accepted or rejected based on consideration of a plurality
of its read properties “to a single comparison.” App. Br. 8–9. Appellants
contend the above-mentioned classification process was “not previously
known . . . as evidenced by the fact that there are not art rejections of the
claims” and that “realize[] an improvement in computer functionality” over
“traditional variant calling approaches, which only recognize the base
quality and read coverage as tools” for discarding potentially erroneous
reads during assembly of the genetic sequence. See id. 15–17. As such,
Appellants argue that to the extent there is a judicial exception recited in the
claims it is “integrated into at least one practical application” and/or the

2 Examiner does not specifically address Guidance Step 2A Prong 2 at step
one of the Alice analysis.
Appeal 2019-002987
Application 14/363,436

8
claims “recite elements that are not well-understood, routine or
conventional.” Id. at 17–18.
Our Review
Applying the Supreme Court’s Alice framework as explained in the
Office’s Guidance, we agree with Examiner that, given their broadest
reasonable interpretation, the claims recite the mental process of comparing
and analyzing genetic sequence data to identify variants (Guidance Step 2A,
Prong 1).3 However, as explained more fully below, we are persuaded by
Appellants’ arguments that the judicial exception is integrated into a
practical application that improves the functioning of genetic sequencing
technology (Guidance Step 2A, Prong 2). Therefore, Appellants’ claims are
patent eligible.
Guidance Step 2A, Prong 1
Claim 1 and the other independent claims recite “identifying a
possible variant in an assembled genetic sequence comprising aligned
sequencing reads;” and then “calling the possible variant.” In this context,
“calling” a variant means that a portion of the assembled sequence data (e.g.,
from a DNA sample taken from a patient) that varies from the expected
sequence is “accepted as an actual variant for clinical purposes.” Spec. 8, ll.
18–20. Accordingly, the above limitations recite the mental process of
comparing and analyzing genetic sequences to determine differences
between the sequence of a sample and a reference. See In re BRCA1, 774

3 It is undisputed that Appellants’ claims are directed to one of the statutory
classes of patentable subject matter, i.e., a method or “process,” recited in
35 U.S.C. § 101. Thus, we begin our analysis at Step 2A, prong 1 of the
Guidance.
Appeal 2019-002987
Application 14/363,436

9
F.3d at 763 (holding that “‘comparing’ and ‘analyzing’ two gene sequences”
to determine differences between them is an “abstract mental process”)
(quoting Assoc. for Molecular Pathology v. Myriad, 689 F.3d 1303, 1334
(Fed. Cir. 2012)).
Moreover, while the Specification discusses “‘next generation’”
sequencing approaches that employ parallel processing techniques that
enhance throughput by orders of magnitude” to assemble and analyze large
amounts of genetic sequence data, claims 1, 14, and 16 are not limited to
such. Spec. 2, ll. 19–21. Given their broadest reasonable interpretation,
these claims encompass the comparison and analysis of shorter sequences
that could be practically performed in a human mind. Thus, the fact that
claim 1 further requires the claimed method to be “performed by an
electronic data processing device” does not remove the exception recited
here from the mental process category. See Guidance at 52 n. 14 (“If a
claim, under its broadest reasonable interpretation, covers performance in
the mind but for the recitation of generic computer components, then it is
still in the mental processes category unless the claim cannot practically be
performed in the mind.”). For these reasons, we conclude that claims 1, 14,
and 16 recite an abstract idea––a mental process.
Guidance Step 2A, Prong 2
Having determined that the claims recite a judicial exception, our
analysis now turns to determining whether there are additional elements that
integrate the judicial exception into a practical application such that the
claim is not directed to the exception itself (Guidance Step 2A, prong 2).
“Integration into a practical application” requires that the claim recite an
additional element or a combination of elements, that when considered
Appeal 2019-002987
Application 14/363,436

10
individually or in combination, “apply, rely on, or use the judicial exception
in a manner that imposes a meaningful limit on the judicial exception, such
that the claim is more than a drafting effort designed to monopolize the
judicial exception.” Guidance at 54. One consideration “indicative that an
additional element (or combination of elements) may have integrated the
exception into a practical application” is where it “reflects an improvement
in the functioning of a computer, or an improvement to other technology or
technical field.” Id. at 55.
Here, we determine that Appellants’ claims recite additional elements
that integrate the judicial exception into a practical application by improving
the functioning of the genetic sequencing apparatus used to call variants.
Specifically, claim 1 recites “computing values of at least one read property
for sequencing reads of the assembled genetic sequence,” calling the
possible variant “conditional upon the computed values . . . satisfying an
acceptance criterion” and using “a classifier operating on the computed
values of the plurality of read properties for the reads that include the
possible variant” to classify the variant as “accepted or rejected” for the
purpose of calling it an actual variant.4 Collectively, these elements recite
an acceptance test that differs from what the Specification refers to as
“typical[]” variant calling “based on nucleotide base quality and coverage”
filtering because it occurs after the identification of possible variants through
the assembly and alignment of the sample sequence and relies on different
inputs (i.e., read properties as opposed to base quality and coverage). See
Spec. 2, ll. 4–18. Indeed, claim 7 further distinguishes these processes by

4 Claims 14 and 16 recite the same limitations. See App. Br. 22–23.
Appeal 2019-002987
Application 14/363,436

11
requiring that the reads that do not satisfy “base quality score and coverage
criteria” be “discarded . . . and . . . not included in the assembled genetic
sequence” (i.e., the typical process) prior to performing the claimed
acceptance test.
Appellants provide evidence-backed argument supporting that the
addition of the claimed acceptance test realizes an improvement in the
function of existing genetic sequencing technology. In particular,
Appellants rely on a study cited in the Specification, demonstrating that “the
variant call error rate may be substantially higher than the base call error rate
predicated” by the base quality score and that a need exists for error
reduction beyond that which can be achieved by filtering out suspect reads
prior to assembling the sequence. See App. Br. at 12–13 (citing Harismendy
et al., “Evaluation of Next Generation Sequencing Platforms for Population
Targeted Sequencing Studies,” Genome Biology Vol. 10:R32 (2009)).
Appellants contend the claimed acceptance test “solves the problem of high
variant call error rate by the recited use of the classifier.” Id. at 13–14
(citing Specification). Examiner offers no evidence to the contrary. Thus,
the record supports Appellants’ argument that the claims recite an
“improvement over traditional variant calling.” Id.at 16–17.
It is the recitation of this improvement to the variant calling operation
in Appellants’ claims that distinguishes them from the claims held to be
patent ineligible in In re BRCA1. Unlike Appellants’ claims here, the “non-
patent ineligible elements” of the claims in In re BRCA1 did “nothing more
than spell out what practitioners already knew––how to compare gene
sequences using routine, ordinary techniques.” Id. at 764. For example,
claim 7 recited that the sequence was “compared by hybridizing a . . . gene
Appeal 2019-002987
Application 14/363,436

12
probe . . . and detecting the presence of the hybridization product.” Id. at
761. It did not, however, recite “any new process for designing or using
probes, primer, or arrays” such that it effected an improvement to genetic
analysis technology or added anything beyond well-understood and
conventional activity to the judicial exception itself. See id. at 764. In
contrast, Appellants’ claims recite specific steps that improve the function of
the computer in the genetic sequencing apparatus by incorporating an
assessment test based on read properties before accepting or rejecting a
possible variant as an actual one. See Finjin, Inc. v. Blue Coat Sys., Inc. 879
F.3d 1299, 1304 (Fed. Cir. 2018) (holding that claims reciting “specific
steps” to effect a “‘behavior-based’ approach to virus scanning” constituted
an “improvement in computer functionality” and were therefore patent
eligible).
For these reasons, we conclude that independent claims 1, 14, and 16
are not directed to an abstract idea. Thus, we reverse Examiner’s rejection
of those claims and need not proceed to Alice step 2 to determine whether
the claims also provide an inventive concept (USPTO Guidance Step 2B).
See Guidance at 56. For the same reasons, we reverse Examiner’s rejection
of dependent claims 7, 15, 20, 22, and 23, which is premised on the same
rationale. See Final Act. 4–8.
SUMMARY
We reverse the rejection of claims 1, 2, 7, 14–16, 20, 22, and 23 under
35 U.S.C. § 101 as being directed to patent ineligible subject matter.

REVERSED