Ex Parte Duffy

Patent Trial and Appeal Board

Jun 11, 2018

13412887 (P.T.A.B. Jun. 11, 2018)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR
13/412,887 03/06/2012 Nigel P. Duffy
54698 7590 06/13/2018
MOSER TABOADA
1030 BROAD STREET
SUITE 203
SHREWSBURY, NJ 07702
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
ATTORNEY DOCKET NO. CONFIRMATION NO.
NUMOllCON 5269
EXAMINER
SKIBINSKY, ANNA
ART UNIT PAPER NUMBER
1631
NOTIFICATION DATE DELIVERY MODE
06/13/2018 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):
docketing@mtiplaw.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte NIGEL P. DUFFY 1
Appeal2017-000697
Application 13/412,887
Technology Center 1600
Before RICHARD M. LEBOVITZ, RICHARD J. SMITH, and
TIMOTHY G. MAJORS, Administrative Patent Judges.
SMITH, Administrative Patent Judge.
DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134 involving claims to a method
for training a molecular properties model and predicting a property of
interest for a test molecule. We have jurisdiction under 35 U.S.C. § 6(b).
We affirm.
STATEMENT OF THE CASE
Background
"In the fields of bioinformatics and computational chemistry, machine
learning applications have been used to develop models of various molecular
1 According to Appellant, the real party in interest is Numerate, Inc.
( Appeal Br. 3.)
Appeal2017-000697
Application 13/412,887
properties. Oftentimes, such models are built in an attempt to predict
whether a particular molecule will exhibit the property being modeled."
(Spec. ,r 5.)
"The training examples used to train a molecular properties model
each typically include a description for a molecule ... and data regarding the
property of interest for the molecule. Collectively, the training examples are
commonly referred to as a 'training set' or as 'training data."' (Id. ,r 6.)
"[I]t is often easier ( and more accurate) to consider the ordering of
two molecules relative to a certain property than it is to assert an absolute
value for the property for a single molecule. Existing molecular property
modeling techniques, however, are not capable of using such ordering
information." (Id. ,r 11.)
Claims on Appeal
Claims 11-23 are on appeal. 2 (Claims Appendix, Appeal Br. 12-14.)
Claim 11 is the only independent claim on appeal and reads as follows:
11. A method for training a molecular properties model and
predicting a property of interest for a test molecule, comprising,
on a computer system:
obtaining a pseudo-partial ordering of molecules represented by
a set of two or more ranked pairs of molecules, wherein
the molecules within the pairs are ordered relative to one
another based on the property of interest that is physically
testable selected from the group consisting of reactivity,
binding affinity, melting point, solubility, membrane
permeability, toxicity, pKa, and combinations thereof;
generating a representation of the molecules included in the
pseudo partial ordering of molecules that is appropriate for
a selected machine learning algorithm;
2 Claims 1-10 and 24--41 are cancelled. (See Final Office Action dated Oct.
19, 2015 ("Final Act."), at 2.)
2
Appeal2017-000697
Application 13/412,887
providing as a training set the set of pairs of the pseudo partial
ordering of molecules to the selected machine learning
algorithm that optimizes a function of the pairs, wherein
the molecules that comprise the pairs are ranked correctly
in the obtained ranked pairs relative to the property of
interest, and wherein, for at least a pair in the training set,
said function penalizes an incorrectly ordered said pair
implied by the function;
executing on the computer system the selected machine learning
algorithm to generate a trained molecular properties
model;
generating a representation of an additional molecule appropriate
for the molecular properties model, which molecule is the
test molecule; and
using the molecular properties model, generating on the
computer system a prediction of a value of the property of
interest for said test molecule; and
further comprising independent of the computer system
determining the accuracy of the prediction for said test
molecule by performing a laboratory confirmation of the
predicted property of interest using a physically realized
sample of said test molecule.
Examiner's Rejection
Claims 11-23 stand rejected under 35 U.S.C. § 101 as directed to non-
statutory subject matter. (Ans. 3--4.)
DISCUSSION
Issue
Whether a preponderance of evidence of record supports the
Examiner's rejection of claims 11-23 under 35 U.S.C. § 101.
Principles of Law
On issues of patent eligibility, we "first determine whether the claims
at issue are directed to a patent-ineligible concept," such as laws of nature,
3
Appeal2017-000697
Application 13/412,887
natural phenomena, and abstract ideas. Alice Corp. Pty. Ltd. v. CLS Bank
Int'!, 134 S. Ct. 2347, 2355 (2014) ("Alice"). If this threshold is met, we
move to the second step of the inquiry and "consider the elements of each
claim both individually and 'as an ordered combination' to determine
whether the additional elements 'transform the nature of the claim' into a
patent-eligible application." Id. (quoting Mayo Collaborative Servs. v.
Prometheus Labs., Inc., 566 U.S. 66, 78-79 (2012) ("Mayo")).
Analysis
We adopt the Examiner's findings and conclusion that the claims are
directed to non-statutory subject matter. (Advisory Action dated Jan. 20,
2016, at 2; Ans. 3-7.) We discern no error in the rejection of claims 11-23
under Section 101. We limit our consideration to claim 11 because the
claims were not argued separately.
According to Appellant, claim 11 is directed to a method of
determining "the properties of a molecule with a machine learning algorithm
that uses ranked pairs of molecules as the training data. Using this relative
ranking allows robust computer models to be constructed in situations where
data sources are too noisy to allow such modeling using traditional
valuations of the molecules." (Appeal Br. 3--4.)
Examiner's Position
As to Alice step one, the Examiner finds that claim 11 is
directed to obtaining a list of ordered molecules, generating a list
of data representing the molecules, providing the "pairs of
molecules" (which is in data form) to an algorithm that ranks the
molecules according to a molecular property of interest,
generating further data representing an additional molecule,
[ and] predicting information regarding the property of interest
4
Appeal2017-000697
Application 13/412,887
for the additional molecule. The method further reqmres a
laboratory "confirmation," of the predicted information about
property of interest in the additional molecule using a physical
sample of the model. As a whole the method relies on the abstract
ideas of comparing information, categorizing, organizing and
transmitting information and organizing information through
mathematical correlations.
(Ans. 3--4.)
Regarding Alice step two, the Examiner finds that claim 11
do [es] not include additional elements that are sufficient to
amount of significantly more than the judicial exception because
it is routine and conventional to perform the acts of testing a
physical molecular sample to determine at least one property of
the molecules. Other elements of the method include using a
computer system to obtain and generate data as well as perform
the algorithmic calculation. However, this is a recitation of
generic computer structure that serves to perform generic
computer functions that are well-understood, routine, and
conventional activities previously known to the pertinent
industry. Viewed as a whole, these additional claim element(s)
do not provide meaningful limitation(s) to transform the abstract
idea into a patent eligible application of the abstract idea such
that the claim(s) amounts to significantly more than the abstract
idea itself.
(Id. at 4.)
Based on those findings, the Examiner concludes that claim 11 is
directed to non-statutory subject matter and unpatentable under Section 101.
(Id.)
Appellant's Arguments
Appellant argues that "nothing in this record indicates that a
molecular properties model based on ordered pairs of molecules is known or
routine. Thus, on this record, and Applicant submits in actual fact, there is
nothing routine or conventional about the claimed method." (Appeal Br. 5.)
5
Appeal2017-000697
Application 13/412,887
We are not persuaded. As an initial matter, the Specification makes
clear that the use of molecular properties models, training sets, and machine
learning algorithms are routine and conventional in the fields of
bioinformatics and computational chemistry. (See, e.g., Spec. ,r,r 5, 6.)
Moreover, the Specification indicates that "any available computer system"
may be used and identifies by name multiple prior learning algorithms that
may be used in the claimed method. (See Spec. ,r,r 29, 83.)
Regarding the use of ranked pairs of molecules ("ordered pairs") as
the training data, the Examiner explains that "the molecular properties model
is one based on mathematical relationships, algorithms and data. It is drawn
to an abstract idea." (Ans. 6.) That is, providing training data in the form of
ordered pairs involves the abstract idea of comparing, organizing, and
transmitting information. 3 (Id. at 4.) We thus agree with the Examiner's
finding because abstract ideas include collecting information and analyzing
that information "by steps people go through in their minds, or by
mathematical algorithms." FairWarning IP, LLC v. Iatric Sys., Inc., 839
F.3d 1089, 1093 (Fed. Cir. 2016) (internal quotation marks and citation
omitted). Put concisely, "[ w ]ithout additional limitations, a process that
employs mathematical algorithms to manipulate existing information to
generate additional information is not patent eligible." Digitech Image
Techs., LLCv. Elecs.for Imaging, Inc., 758 F.3d 1344, 1351 (Fed. Cir.
2014). 4
3 A model based on ordered pairs of molecules is thus an abstract idea under
Alice step 1. (Ans. 3--4, 6.)
4 See also Intellectual Ventures I LLC v. Capital One Fin. Corp., 850 F.3d
1332, 1340 (Fed. Cir. 2017), discussing cases in which claims reciting data
manipulation steps were held to be patent ineligible as abstract ideas.
6
Appeal2017-000697
Application 13/412,887
Appellant argues that the claimed invention improves the functioning
of the computer. (Appeal Br. 6-8.) In advancing that argument, Appellant
characterizes the invention as a form of digital image processing of the type
that was found patent eligible in Research Corp. Techs., Inc. v. Microsoft
Corp., 627 F.3d 859 (Fed. Cir. 2010). 5 (Id. at 6.) Appellant also relies on
the Supplemental Allgood Declaration6 to support the contention that the
claimed invention improves the functioning of the computer. (Id. at 7.)
According to Appellant, "the claimed method using ordered pairs improves
the method of visualizing the activity of molecules." (Id.)
We are not persuaded. As explained by our reviewing court in
Bancorp Servs., L.L.C. v. Sun Life Assur. Co. of Canada (U.S.), 687 F.3d
1266, 1279 (Fed. Cir. 2012), "the claimed processes in Research Corp.
plainly represented improvements to computer technologies in the
marketplace," such as the use of an inventive mask that produced "higher
quality halftone images while using less processor power and memory
space." (Id., quoting Research Corp., 627 F.3d at 865.) As further
explained in Bancorp, the method in Research Corp. "required the
manipulation of computer data structures ... and the output of a modified
computer data structure ... dependent upon the computer components
required to perform it." Id.
5 Appellant references Example 3 of EXAMPLES: ABSTRACT IDEAS issued by
the USPTO on January 27, 2015, rather than Research Corp., but that
Example 3 was based on the Research Corp. decision. See id. at 7-10
(https ://www. uspto. gov/ sites/ default/files/ documents/ abstract_idea_example
s.pdf).
6 Supplemental Declaration Under 37 C.F.R. § 1.132 of Brandon Allgood,
dated Dec. 18, 2015 ("Allgood Declaration" or "Deel.").
7
Appeal2017-000697
Application 13/412,887
Here, the focus of claim 11 is not to any improved computer, but
rather to an improved mathematical analysis, and the Specification states
that readily available computer technology is usable to carry out the
analysis. See SAP Am., Inc. v. Investpic, LLC, 890 F.3d 1016, 1023 (Fed.
Cir. 2018). Stated another way, claim 11 is not directed to a physical-realm
improvement, "but an improvement in wholly abstract ideas-the selection
and mathematical analysis of information, followed by reporting or display
of the results." Id. at 1022. Moreover, given Appellant's contention that
"[t]he invention is to a method of determining the properties of a molecule
with a machine learning algorithm that uses ranked pairs of molecules as the
training data" (Appeal Br. 4), we find that the recited generic computer
elements "'are invoked merely as a tool."' Credit Acceptance Corp. v.
Westlake Servs., 859 F.3d 1044, 1055 (Fed. Cir. 2017) (quoting Enfish, LLC
v. Microsoft Corp., 822 F.3d 1327, 1335-36 (Fed. Cir. 2016)).
The Allgood Declaration does not change this analysis. Rather than
describing an actual improvement to a computer itself, the Allgood
Declaration states that the use of ordered pairs of molecules "provides for
more robust input, resulting in better models and better predictions." (Deel.
,r 4.) That statement does not describe an improvement to the computer
itself, but (at most) an improvement in wholly abstract ideas. See SAP, 890
F.3d at 1022.
Appellant argues that "these predictions with ordered pairs is [not] in
the prior art," "an important invention is achieved" by the claims, and that
"[ o ]ne is not monopolizing the field of computational chemistry, as
conventional methods work." (Appeal Br. 7-8.) We are not persuaded. It is
not enough for subject-matter eligibility that a claimed method be novel and
8
Appeal2017-000697
Application 13/412,887
nonobvious in light of prior art. See SAP, 890 F.3d at 1018; Synopsys, Inc.
v. Mentor Graphics Corp., 839 F.3d 1138, 1151 (Fed. Cir. 2016) ("a claim
for a new abstract idea is still an abstract idea"). Moreover, even if the
claimed method is "brilliant," "innovative," or "groundbreaking," that is not
enough for eligibility. See SAP, 890 F.3d at 1018 (internal citations,
quotations, and alteration omitted). Furthermore, the alleged lack of
"monopolizing the field" is unpersuasive because "the absence of complete
preemption does not demonstrate patent eligibility." Ariosa Diagnostics,
Inc. v. Sequenom, Inc., 788 F.3d 1371, 1379 (Fed. Cir. 2015).
Appellant also relies on DDR Holdings, LLC v. Hotels.com, L.P., 773
F.3d 1245, 1257 (Fed. Cir. 2014), and the argument that the solution of
claim 11 is "necessarily rooted in computer technology." (See Appeal Br.
10, quoting DDR Holdings.) However, as explained by our reviewing court,
the focus of the claims in DDR Holdings was to an improvement to
computer functionality itself, rather than simply the use of computers.
Affinity Labs of Tex., LLC v. DIRECTV, LLC, 838 F.3d 1253, 1261---62 (Fed.
Cir. 2016) (distinguishing DDR Holdings). Moreover, rather than being
rooted in computer technology, the claims at issue here simply use computer
technology "as tools in the aid of a process focused on an abstract idea." Id.
at 1262.
Appellant's Reply Brief cites to Enfzsh and McRO, Inc. v. Bandai
Namco Games Am. Inc., 837 F.3d 1299 (Fed. Cir. 2016) as support for the
proposition that "when the invention improves the operation of a computer
or relevant technology with a specific means or method, you do not even
have an abstract idea exception for which you need to seek 'something
more."' (Reply Br. 3.) Even if we accept Appellant's characterization of
9
Appeal2017-000697
Application 13/412,887
McRO and Enfzsh, claim 11 is not directed to the improved operation of a
computer but to a claimed improvement to "the selection and mathematical
analysis of information, followed by reporting or display of the results." See
SAP, 890 F.3d at 1022 (distinguishing McRO and Enfzsh); see also Elec.
Power Grp., LLC v. Alstom S.A., 830 F.3d 1350, 1353-54 (Fed. Cir. 2016)
("collecting information, analyzing it, and displaying certain results of the
collection and analysis" "fall[ s] into a familiar class of claims 'directed to' a
patent-ineligible concept," that of the abstract idea) (distinguishing claims in
Enfish).
Accordingly, for the reasons of record and as set forth above, we
affirm the rejection of claim 11 as being directed to a judicial exception
without significantly more, and thereby being directed to non-statutory
subject matter under 35 U.S.C. § 101. Claims 12-23 fall with claim 11
because they were not argued separately.
Conclusion
A preponderance of evidence of record supports the Examiner's
rejection of claims 11-23 under 35 U.S.C. § 101.
SUMMARY
We affirm the rejection of all claims on appeal.
TIME PERIOD FOR RESPONSE
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a).
AFFIRMED
10