DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
The claim set received on 28 November 2023 has been entered into application.
Claims 1-10, 12-34 and 36-44 are pending.
Priority
This Application is a Continuation of PCT/US2023/030635 filed 18 August 2023 and claims benefit to US Provisional Application 63/371,983 19 August 2022.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 02 December 2023 and 21 July 2026 have been considered by the examiner.
Drawings
The drawings were received on 18 August 2026. These drawings are accepted.
Specification
The Specification received 18 August 2023 has been entered into application.
The disclosure is objected to because it contains an embedded hyperlink and/or other form of browser-executable code. Applicant is required to delete the embedded hyperlink and/or other form of browser-executable code; references to websites should be limited to the top-level domain name without any prefix such as http:// or other browser-executable code. See MPEP § 608.01.
The executable code is in the specification [page 11 para 0067]. The executable code is “https://blast.ncbi.nlm.nih.gov” . It is recommended to amend the specification to recite “ blast.ncbi.nlm.nihdotgov or “blast/ncbi/nlm/nih” . It is noted the “https://” and “.gov” are to be cancelled from the specification.
Claim Objections
Claim 42 is objected to because of the following informalities: Claim 42 recites “computer program code that, when executed by a computing system, cause the computing system to perform operations including”. The claim should be amended to recite “computer program code that, when executed by a computing system, causes the computing system to perform operations including”. Appropriate correction is required to address the grammatical correctness of the claim.
Claim Rejections - 35 USC § 112
35 USC § 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 3-7, 13, and 42 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 3 recites “chemical mapping data is generated by a process comprising contacting the RNA molecule with a chemical probing agent” while claims 4-7 provide the chemical probing agents. Here, it is not clear how the physical chemical probes are to limit claim 2 when claim 2 is drawn to a computer implemented method that does not contain any physical chemical probing steps. It is recommended to amend claim 2 to encompass a data gathering step(s) (i.e., chemical probing step) and cancel the “computer implemented method” in the preamble to provide a method that can incorporate and analyze the chemical probing data for subsequent RNA tertiary structure prediction such as “A method of predicting tertiary structure of an RNA molecule of interest comprising: obtaining chemical mapping data using chemical probing methods comprising SHAPE and/or DMS probes…”, for example.
Claim 42 recites “accessing an RNA tertiary structure prediction system that was manufactured by a process comprising”. The limitation renders the claim indefinite because it is not clear if a computer or CRM is performing the accessing, a computer or CRM. Here, it is recommended to amend the claims to recite “ One or more non-transitory computer-readable medium (CRM) comprising instructions for: accessing an RNA tertiary structure prediction system that was manufactured by a process comprising”.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefore, subject to the conditions and requirements of this title.
Claims 1-10, 12-34 and 36-44 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to non-statutory subject matter.
Step I - Process, Machine, Manufacture or Composition
Claim 1 is drawn to a method, so a process.
Claim 2-10, 12-34, 36-39, and 44 are drawn to a method, so a process.
Claim 40 is drawn to a system, so a machine.
Claim 41 is drawn to a computer-readable medium (CRM), so a manufacture.
Claim 42 is drawn to a computer-readable medium (CRM), so a manufacture.
Claim 43 is drawn to a method, so a process.
Step 2A Prong I - Identification of an Abstract Idea
Claims 1 and 2 recite a computer-implemented method and claim 40 recites a computing device while claim 41 recites using computer modules. Claims 1-2 and 40-43 recite similar limitations and claim language and are therefore examined similarly. Claim 42 contains an accessing an RNA tertiary prediction system step. This step of claim 42, however, is examined individually but in conjunction with the limitations of claims 1-2, 40-41, and 43.
Claims 1-2 and 40-43 recite:
Claims 1 (creating), 2 (a)(i), 40(a), 41(a), 42(accessing an RNA tertiary prediction system), and 43 (wherein the machine-learning model was trained by a process including (i) recite:
creating a training data set comprising chemical mapping data for a first plurality of RNA molecules and tertiary structure data for a second plurality of RNA molecules
This step can be performed in the human mind by organizing data (i.e., chemical mapping data and tertiary structure data) to create organized information (training data set) and is therefore an abstract idea.
Claims 1 (training), 2(a)(ii), 40(b), 41(b), 42 (training), and 43 (a)(ii) recite:
training a machine learning algorithm using the training data set
This step encompasses taking information (i.e., dataset: chemical mapping data and tertiary structure data) and manipulating the data via mathematical correlation (i.e., machine learning algorithm) for organizing the data into a different form (i.e., trained dataset) which reads on abstract ideas. See MPEP 2106.04(a)(2)(III)(A)(iv) and 2106.04(a)(2)(III)(C)(1-3). Here, the training step is broadly and generically recited and reads on abstract ideas.
Claims 1 (applying), 2 (b), 40(c), 41(c), 42 (using step), and 43 (wherein the machine learning algorithm generates the tertiary structure) recite:
applying the trained machine learning algorithm to predict the tertiary structure of the RNA molecule of interest
This step encompasses taking information (i.e., trained dataset of chemical mapping data and tertiary structure data) and manipulating the data via mathematical correlation (i.e., machine learning algorithm) for organizing the data into a different form (i.e., to predict the tertiary structure of the RNA molecule of interest) which reads on abstract ideas. See MPEP 2106.04(a)(2)(III)(A)(iv) and 2106.04(a)(2)(III)(C)(1-3). Here, the applying step is broadly and generically recited and reads on abstract ideas.
Claim 2 also recites:
Claim 2 recites:
(a) obtaining a machine-learning model
This encompasses obtaining a machine leaning model which encompasses mathematical formulas and mathematical steps (i.e., algorithms) for organizing and manipulating data for training and applying a machine model for subsequently predicting RNA tertiary structure which reads on abstract ideas (i.e., mathematical concepts/mathematical formulas). See MPEP 2106.04(a)(2)(I)(A-C). It is noted that while “machine-learning model” does not recite specific mathematical formulas, formulas, and mathematical steps (i.e., algorithm) machine-learning models by nature encompass mathematics and/or statistics (e.g., vector mathematics [Spec page 37 para 0153]), either explicitly or implicitly. Furthermore, the obtaining step is broadly and generically recited and reads on abstract ideas (i.e., obtaining a mathematical model).
Dependent claim(s):
Claims 8-10, 12-34, 37-39, and 44 are further drawn to limitations that describe the abstract ideas of claim 1 and are therefore also abstract ideas.
Step 2A Prong II - Consideration of Practical Application
Here, in the instant case, claims 1-2, 40-42 and 43 merely set forth methods of data analysis for outputting a predicted RNA tertiary structure (claims 1-2, 40-42) and receiving RNA molecule data (claim 43). It is noted that outputting and receiving data are deemed extra-solution activities that do not integrate the judicial exception into a practical application. As such, practicing the claims merely results in outputting a predicted RNA structure of molecules and receiving RNA molecule data (claim 43). Such a result only produces information and does not provide for a practical application in the physical-realm of physical things and acts, i.e., the claims do not utilize the data generated by the judicial exception to affect any type of change. See MPEP 2106.04(a)(2)(A)(iv).
Claims 1-2, 40-42 and 43 recites “training a machine learning algorithm using the training data set” and “applying the trained machine learning algorithm to predict the tertiary structure of the RNA molecule of interest”. Here, even though the claimed steps utilize a “machine learning algorithm (MLA)”, the MLA is broadly and generically recited and, as noted in Step 2A Prong II above, reads on mere instructions to implement an abstract idea on a generic computer in a computer environment and reads on mathematical/statistical computations. See 2024 Subject Matter Eligibility Update (AI) [Example 47 Claim 2] and MPEP 2106.04(a)(2)(III)(C)(1-3) and 2106.05(f).
Furthermore, and for sake of compact prosecution, even if the MLA of claims 1-2 and 40-43 is also considered an additional element, the MLA is used to generally apply the abstract idea without limiting how the trained MLA functions. The MLA is described at a high level such that it amounts to using a computer with a generic MLA to apply the abstract idea. These limitations only recite the outcomes for “training a MLA” and “applying the MLA to predict tertiary RNA” without any details about how MLA is trained and without any details about how the MLA is applied for predicting RNA tertiary for outputting predicted tertiary structure which does not integrate the judicial exception into a practical application or provide significantly more because this type of recitation is equivalent to the words "apply it". See 2024 Subject Matter Eligibility Update (AI) [Example 47 Claim 2] and MPEP 2106.05(f).
Claims 1-2, 40-42 and 43 recite “creating a training data set…”. These steps read on the abstract ideas of organizing data (i.e., chemical mapping data and tertiary structure data) to create organized information (training dataset), not physical creating a tangible object. Therefore, the creating step does not integrate the recited judicial exception into a practical application.
Claim 14 recites “further comprising, before applying the machine-learning model, tuning the machine learning algorithm based on a chemical mapping data of the RNA molecule of interest.” This step encompasses performing mathematical computations for refining/tuning the MLA based on chemical mapping data which reads on abstract ideas. Here, even though the claimed steps tune a “machine learning algorithm (MLA)”, the MLA is broadly and generically recited and, as noted in Step 2A Prong II above, reads on mere instructions to implement an abstract idea on a generic computer in a computer environment and reads on mathematical/statistical computations. See 2024 Subject Matter Eligibility Update (AI) [Example 47 Claim 2] and MPEP 2106.04(a)(2)(III)(C)(1-3) and 2106.05(f).
Claim 32 recites “further comprising formulating a pharmaceutical drug based on the predicted tertiary structure”. This recitation of the claimed limitations “formulating a pharmaceutical drug “ that attempts to cover any solution for formulating a pharmaceutical drug with no restriction on how “formulating a pharmaceutical drug” is accomplished and no description of the mechanism for accomplishing formulating a pharmaceutical drug which does not integrate a judicial exception into a practical application or provide significantly more because this type of recitation is equivalent to the words "apply it". See MPEP 2106.05(f)(1).
This judicial exception is not integrated into a practical application because the claims do not meet any of the following criteria:
an additional element reflects an improvement in the functioning of a computer, or an improvement to other technology or technical field;
an additional element that applies or uses a judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition;
an additional element implements a judicial exception with, or uses a judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim;
an additional element effects a transformation or reduction of a particular article to a different state or thing; and
an additional element applies or uses the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception.
Step 2B: Consideration of Additional Elements and Significantly More
The claimed method also recites "additional elements" that are not limitations drawn to an abstract idea.
The recited additional elements of using computer processes, components, and equipment of claims 1-2, 40-43 does not add significantly more than the recited judicial exception because using computer element to process, store, and evaluate RNA data in a computing environment is merely tangential to the claimed method and is deemed conventional. MPEP 2105.06(d)(II) and 2106.05(g).
To provide conventionality of using computer processes for predicting Yu et al. (Yu) teaches different neural networks and machine learning processes for predicting RNA structures/tertiary structures [page 2 fig 1, page 3 table 1] (Frontiers in molecular biosciences, 2022-05, Vol.9, p.869601) (Cited in IDS received 21 July 2026, NPL Cite No.: 004).
The recited additional elements of using computer processes, components, and equipment of claims 1-2 (computer implemented method), 40 (computer device), 41-42 (CRM), and 43 (computer implemented method) does not add significantly more than the recited judicial exception because using computer element to process, store, and evaluate RNA data in a computing environment is merely tangential to the claimed method and is deemed conventional. MPEP 2105.06(d)(II).
The recited additional elements of data output of claims 1 (last step), 2 (c), 40 (d), 41(d), and 42 (last step) does not add significantly more than the recited judicial exception because outputting data to display results is deemed a routine and conventional extra-solution activity. See MPEP 2106.05(g).
The recited additional elements of receiving data of claims 43 step (b) (last step) does not add significantly more than the recited judicial exception because receiving data that is subsequently evaluated by the abstract ideas is deemed a routine and conventional extra-solution activity. See MPEP 2106.05(g).
The recited additional elements of data output of claims 1 (last step), 2 (c), 40 (d), 41(d), and 42 (last step) does not add significantly more than the recited judicial exception because outputting data to display results is deemed a routine and conventional extra-solution activity. See MPEP 2106.05(g).
The recited additional elements of using computer processes, components, and equipment for storing data of claim 42 does not add significantly more than the recited judicial exception because using computer elements to store algorithms is merely tangential to the claimed method and is deemed conventional. MPEP 2105.06(d)(II) and 2106.05(g).
The recited additional elements of using chemical probing agents of claims 3-7 do not add significantly more than the recited judicial exception because using chemical probes as reagents data is a deemed conventional extra-solution activity. MPEP 2105.06(d)(II) and 2106.05(g).
To provide evidence of conventionality, Busan et al. (Busan) teaches RNA probing and structure analysis and modeling investigations focusing on in vitro transcribed and cell-free extracted RNAs, SHAPE experiments using 1M7 modification and MaP readout broadly satisfy these criteria. Busan also teaches that 1M6, 1M7, NMIA, and 5NIA are used as probing agent/contacting agents [page 9 discussion].
Busan also teaches evaluating RNA structure using SHAPE probing agents (1M7, 1M6, and NMIA), one more recently proposed (NAI), and one novel reagent (5NIA) [abstract]. Busan teaches using SHAPE reagents (i.e., probing agents) DMS modified RNA [page 8] and teaches reagents utilized 1M7, 1M6, NMIA, 5NIA, and NIA [page 21 table 1] (Biochemistry (Easton), 2019-06, Vol.58 (23), p.2655-2664).
To provide conventionality of using SHAPE reagent 2A3, Marinus et al. (Marinus) teaches using SHAPE reagents (i.e., probes) for evaluating RNA structure [abstract] (Biochemistry (Easton), 2019-06, Vol.58 (23), p.2655-2664) (Nucleic acids research, 2021-04, Vol.49 (6), p.e34-e34).
The recited additional elements of formulating a drug of claim 32 does not add significantly more than the recited judicial exception because formulating drugs from RNA is deemed conventional. MPEP 2105.06(d)(II) and 2106.05(g).
The recited additional elements of inputting by sending a query to a query of claim 43 (a) does not add significantly more than the recited judicial exception because providing input (i.e., query) is deemed a conventional extra solution activity. MPEP 2105.06(d)(II) and 2106.05(g).
In conclusion, and when viewed as a whole, these additional claim element(s) do not provide meaningful limitation(s) to transform the abstract idea recited in the instantly presented claims into a patent eligible application of the abstract idea such that the claim(s) amounts to significantly more than the abstract idea itself. Therefore, the claim(s) are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1-5, 8-10, 13-15, 20-25, 28-29, 33-34, and 36-43 are rejected under 35 U.S.C. 103 as being unpatentable over Pearce et al. (bioRxiv, 2022-05) in view of Wei et al. (Briefings in Bioinformatics, 2022-01, Vol.23 (1), p.bbab540).
Claims 1 and 2 recite a computer-implemented method and claim 40 recites a computing device while claim 41 recites using computer modules. Claims 1-2 and 40-41 recite similar limitations and claim language and are therefore examined similarly. Claim 42 contains an accessing an RNA tertiary prediction system and storing the trained machine learning step while claim 43 contains a sending query step. These steps of claim 42 and 43, however, are examined individually but in conjunction with the limitation similar to claims 1-2 and 40-41.
Claims 1-2 and 40-43 recite:
creating a training data set comprising chemical mapping data for a first plurality of RNA molecules and tertiary structure data for a second plurality of RNA molecules ,claims 1 (creating), 2 (a)(i), 40(a), 41(a), 42(accessing an RNA tertiary prediction system), and 43 (wherein the machine-learning model was trained by a process including (i).
training a machine learning algorithm using the training data set, claims 1 (training), 2(a)(ii), 40(b), 41(b), 42 (training), and 43 (a)(ii).
applying the trained machine learning algorithm to predict the tertiary structure of the RNA molecule of interest, claims 1 (applying), 2 (b), 40(c), 41(c), 42 (using step), and 43 (wherein the machine learning algorithm generates the tertiary structure).
Claims 1 (outputting), 2 (c), 40(d), 41(c), and 42 (outputting step) recite outputting the predicted tertiary structure of the RNA molecule of interest, claims 1 (outputting), 2 (c), 40(d), 41(c), and 42 (outputting step).
Claims 2 and 42-43 also recite:
Claim 2:
Claim 2 recites (a) obtaining a machine-learning model.
Claim 42:
Claim 42 recites accessing an RNA tertiary structure prediction system that was manufactured by a process comprising:
Claim 42 recites storing the trained machine-learning model on the non-transitory computer-readable medium.
Claim 43
Claim 43 recites (a) sending a query for predicting the tertiary structure of the RNA molecule of interest to a computer comprising a machine-learning model.
Claim 43 (b) recite receiving the predicted tertiary structure of the RNA molecule of interest from the computer.
Pearce
Pearce et al. (Pearce) teaches using pairwise distance maps between atoms N1/N19 atoms, C4’ atoms, and backbone P atoms, and the inter-residue and backbone torsion angles [Pearce, page 10 geometric restraint prediction]. Pearce teaches analyzing C3’ atoms [Pearce, page 19] [Spec page 15 para 0079]. Pearch teaches using RNA-puzzles targets/structures (i.e., Puzzle 5 is a 188-nucleotide long lariat-capping ribozyme (PDB ID: 4p9r), Puzzle 6 is a 168-nucleotide adenosylcobalamin riboswitch (PDB ID: 4gxy), Puzzle 7 is the Varkud satellite ribozyme (PDB ID: 4r4v), and Puzzle 12 is a medium-size (108 nucleotides) ydaO riboswitch (PDB ID: 4qlm) which are RNA tertiary structures [Pearce pages 5-6], as in claims 1 (creating…tertiary structure data…), 2 (a)(i)(… tertiary structure data…), 40(a)(…tertiary structure data…), 41(a), 42(accessing an RNA tertiary prediction system… tertiary structure data…), and 43 (wherein the machine-learning model was trained by a process including (i) (…tertiary structure data).
Pearce teach DeepFoldRNA pipeline using neural networks [page 16 Figure 1]. Pearce teaches DeepFoldRNA (i.e., neural network algorithm) was trained using 2,986 RNA chains collected from ProteinDataBase (PDB) [Pearce, page 10 training data and Procedure], as in the training step of claims 1 (training), 2(a)(ii), 40(b), 41(b), 42 (training), and 43 (a)(ii) training a machine learning algorithm using the training data set. Here, although Pearce doesn’t explicitly teach a “training step” but because Pearce teaches a using a neural network (i.e., DeepFoldRNA) for simulating/modeling the models of the Puzzle Structures from the RNA Puzzle community target/structures (i.e., Puzzle 12 is a medium-size (108 nucleotides) ydaO riboswitch (PDB ID: 4qlm) [Pearce pages 5-6]), it is obvious that DeepFoldRNA would encompass and/or require step(s) for training DeepFoldRNA using training data.
Pearce teaches DeepFoldRNA is a deep learning-based approach to full-length RNA structure modeling, which consists of three main steps: input feature generation, spatial restraint prediction, and L-BFGS folding simulations, as depicted in Figure 1 [Pearce, page 8 methods]. Pearce teaches DeepFoldRNA pipeline using neural networks. Pearce using the model to output a final model [page 16 Figure 1]. Pearce teaches outputting different predicted tertiary/3D RNA structure [Pearce pages 18 fig 3 and page 20 fig 5], as in claims 1 (applying), 2 (b), 40(c), 41(c), 42 (using step), and 43 (wherein the machine learning algorithm generates the tertiary structure) applying the trained machine learning algorithm to predict the tertiary structure of the RNA molecule of interest and claims 1 (outputting), 2 (c), 40(d), 41(c), and 42 (outputting step) outputting the predicted tertiary structure of the RNA molecule of interest
Claim 2
Pearce teaches using DeepFoldRNA which is a neural network pipeline what contains neural network models/algorithm [Pearce, page 16 Fig 1], as in claim 2 (a) obtaining a machine-learning model.
Claim 42:
Pearce teaches using DeepFoldRNA for predicting tertiary RNA structure [Pearce, page 16 Fig 1], as in claim 42 accessing an RNA tertiary structure prediction system that was manufactured by a process comprising.
Pearce teaches each component of the program, including the deep learning models and L-BFGS optimization pipeline, are integrated into a stand-alone package at https://github.com/robpearc/DeepFoldRNA and an online webserver is available at https://zhanggroup.org/DeepFoldRNA, from which users can generate structure models for their own RNA of interest [Pearce page 2], as in claim 42 storing the trained machine-learning model on the non-transitory computer-readable medium. Here, because the DeepFoldRNA software is available both as a stand-alone and as an online webserver, it is obvious the programs are and can be stored on a computer or CRM.
Claim 43
Pearce teaches sending a query sequence into a neural network model for predicting a final 3D structure model (i.e., tertiary structure) [Pearce, page 16 fig 1], as in claim 43 (a) sending a query for predicting the tertiary structure of the RNA molecule of interest to a computer comprising a machine-learning model.
Pearce teaches a final model [Pearce, page 16 fig 1]. Pearce teaches other models generated by DeepFoldRNA [Pearce, page 18 fig 3 and page 20 fig 5], as in claim 43 (b) receiving the predicted tertiary structure of the RNA molecule of interest from the computer.
Dependent claim(s): 10, 15, 20-25, 29, 33-34, and 36-39
Pearce teaches using 2,986 RNA chains [page 10 training and procedure], as in claim 10.
Pearce teaches using DeepFoldRNA to predict RNA structures which uses coupled deep self-attention neural networks with gradient-based folding simulations [Pearce, abstract], as in claim 15.
Pearce teaches using coordinate (i.e., N1/N9, C4, C4) and summary of structure modeling results using DeepFoldRNA [Pearce, page 15 tables 1 and 2]. Pearce teaches the distance maps include the pairwise distances between the nitrogen atoms of the base bonded to the ribose sugar (N1 for pyrimidines and N9 for purines) as well as the backbone C4’ and P atoms [Pearce, page 3 DeepFoldRNA accurately predicts geometric restraints], as in claims 20-23.
Pearce teaches using distance map [Pearce, page 16 fig 1], as in claim 24
Pearce teaches using distance map and torsion angles [Pearce, page 16 fig 1]. [Pearce, page 15 tables 1 and 2], as in claim 25.
Pearce teaches RNA puzzle model was produced with the assistance of fast-track experimental SHAPE data and multidimensional chemical mapping [Pearce page 6]. Pearce teaches DeepFoldRNA with an RMSD of 2.38 Å and a TM-score of 0.796 to the experimental structure, corresponding to a 129.4% improvement in the TM-score over the best model produced during the RNA-Puzzles challenge [Pearce page 6], as in claim 29. Pearce teaches that DeepFoldRNA used the same data as Puzzle 12, and DeepFoldRNA resulted in a 129.4% improvement in the TM-score over the best model produced during the RNA-Puzzles challenge. Here, because Pearce teaches analyzing the experimental structures as the RNA Puzzle challenges, it is obvious Pearce’s DeepfoldRNA uses multidimensional chemical mapping data for one or more RNA molecules of the first plurality of RNA molecules.
Pearce teaches using multiple sequence alignment (MSA) in their neural network [Pearce page 16 fig 1]. Pearce teaches different predicted RNA structures with their associated native structure [Pearce page 18 fig 3, page 20 fig 5], as in claims 33-34. Here, it is obvious that a third molecule or set of molecules would require MSA for prediction.
Pearce teaches an overview of the DeepFoldRNA pipeline which starts from a nucleic acid sequence, multiple RNA sequence databases are searched to create a multiple sequence alignment (MSA) for the query RNA [Pearce, page 16 fig 1]. Pearce teaches producing multiple models [Pearce page 18 fig 3 and 20 fig 5], as in claims 36-38. Here, because Pearce teaches various constructed models, it is obvious that MSA was performed on a third set of RNA molecules, and it is obvious molecules of a first and second RNA molecules are different as Pearce teaches modeling different unrelated RNA structures.
Pearce teaches obtained 105 non-redundant RNA structures from 32 Rfam families after using a sequence identity cutoff of 80% [Pearce, page 3], as in claim 39.
Pearce does not teach:
Pearce does not teach creating a training data set comprising chemical mapping data for a first plurality of RNA molecules of claims 1 (creating), 2 (a)(i), 40(a), 41(a), 42(accessing an RNA tertiary prediction system…comprising…creating a machine learning step), and 43 (wherein the machine-learning model was trained by a process including (i).
Pearch does not teach claims 3-5, 8-9, 13-14, and 28
Wei et al. (Wei) teaches to improve the prediction performance, there are datasets containing experimentally determined RNA structures and in vivo profiles (i.e., chemical mapping data) [Wei page 6 right col RNA secondary structure]. Wei teaches databases that contain secondary RNA structures from transcriptome-wide RNA secondary structure that contain experimentally determined RNA structures (i.e., RNA structures experimented on using chemical mapping methods (i.e., SHAPE-Seq, DMS-seq)) [Wei page 6 right col RNA secondary structure and page 7 table 2]. Wei teaches different data (i.e., in vivo/in vitro structure profiles, secondary, and tertiary structure data) can be used to study interactions between two molecules [Wei page 9 fig 3]. Wei teaches the training data are a set of RNA sequences or RNA secondary structures, which are proved to interact with a protein. Wei teaches a machine learning model or a statistical motif model will be constructed based on the data. Wei teaches the inputs of these models are RNA features, and the models will predict whether they can interact with the protein [Wei page 5 left col Binding preference prediction], as in the creating the dataset elements of claims 1 (creating step), 2 (a)(i), 40(a), 41(a), 42(“accessing an RNA tertiary prediction system…creating a training dataset…”), and 43 (wherein the machine-learning model was trained by a process including: (i) creating a training…”).
Here, Wei teaches machine learning algorithms that can utilize a combination(s) of experimentally determined RNA structure data (i.e., chemical mapping methods (i.e., SHAPE-Seq, DMS-seq)) and other structure data (i.e., in vivo/in vitro structure profiles, secondary, and tertiary structure data) to predict RNA models (i.e., tertiary structure) that interact with proteins. Therefore, Wei teaches a using/creating a dataset of RNA tertiary structure data and chemical mapping/experimental structure data.
Dependent claim(s): 4-5, 8-9, 13-14, and 28
Wei teaches using secondary structure chemical mapping datasets from SHAPE-Seq and DMS-seq data [Wei, page 6 right col RNA secondary structure], as in claim 4. Here, because Wei teaches using DMS-seq datasets, it is obvious that the chemical probing agent DMS was utilized.
Wei teaches using secondary structure chemical mapping SHAPE-Seq datasets [Wei, page 6 right col RNA secondary structure], as in claim 5. Here, because Wei teaches using SHAPE-Seq datasets, it is obvious that the chemical probing agent SHPAE probing were utilized.
Wei teaches using data from RNA Atlas of Structure Probing (RASP) which collects transcriptome-wide RNA secondary structure probing data of 18 experimental methods (i.e., SHAPE-Seq, SHAPE-MaP, and DMS-seq) [Wei, page 6 right col RNA secondary structure]. Wei teaches clean reads are mapped to human genome using STAR [Wei, page 8 top of left col], as in claims 8-9. Here, because Wei teaches using the RASP database, it reads on human transcriptome as RASP encompasses curated multiple high-throughput transcriptome-wide probing experiments involving human cell lines.
Wei teaches using SHAPE and DMS chemical mapping datasets [Wei, page 6 right col RNA secondary structure]. Wei teaches RNA structure features can come from in vivo and in vitro profiling sources [Wei page 9 fig 3], as in claim 13.
Wei teaches the pre-train each module separately first and then fine-tuned all the modules together in an end-to-end fashion [Wei page 12]. Wei teaches RNA structure encodings using icSHAPE data [Wei, page 7 RNA structure encoding], as in claim 14.
Wei teaches typical tertiary RNA structures used in RNA structure prediction (i.e., stems, hairpin loops, pseudoknots) [Wei, page 9 fig 3 ]. Wei teaches using bpRNA, a novel annotation tool to parse complex pseudoknot-containing RNAs with 7 annotations, such as stems, internal loops, bulges, multi-branched loops, external loops, hairpin loops, and pseudoknots [Wei, page 6 right col], as in claim 28.
Obvious claim(s): 3
Wei teaches using secondary structure chemical mapping datasets from SHAPE-Seq and DMS-seq data (i.e., chemical mapping data is generated by a process comprising contacting the RNA molecule with a chemical probing agent) [Wei, page 6 right col RNA secondary structure]. Pearce teaches analyzing native structures and modeled structure which shows different sets of molecules [Pearce pages 18 and 20], as in claim 3.
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Pearce in view of Wei because Wei reviews methods for analyzing RNA structures using machine learning methods trained on experimental (i.e., chemical mapping/probing) data and well-annotated RNA secondary structure data for predicting tertiary structure [Wei page 8 left col]. One of ordinary skill in the art would recognize that while Wei discusses protein and RNA binding Wei also provides a method for determining RNA tertiary structure using JAR3D to calculate the probabilities of an RNA structure folding into the predefined tertiary structural motifs in addition to combining experimental/chemical mapping data (i.e., SHAPE-Seq, DMS-seq, RASP) and RNA structure data (i.e., secondary/tertiary) into input for modeling (i.e., predicting/outputting RNA structure) [Wei, page 6 right col RNA secondary structure]. Here, one of ordinary skill in the art would be motivated to combine Pearce in view of Wei because Wei reviews different methods such as machine learning methods [Wei page 3 fig 2] and using RNA sequencing and RNA structure feature data for determining interactions between proteins and RNA, and Wei teaches secondary structural data (i.e., SHAPE-Seq, DMS-seq, RASP (i.e., chemical mapping data)). Therefore, there is a reasonable expectation of success combining Pearce in view of Wei because Wei teaches combining different RNA feature types such as secondary structure data, experimental chemical probing data (i.e., SHAPE-Seq and DMS-seq (i.e., chemical mapping data (i.e., experimental secondary structural data))), and tertiary structure data [Wei, page 9 fig 3] which highlights the importance of combining structural data information for determining RNA and proteins interaction which can be further utilized for predicting RNA structures that bind to said proteins. Therefore, combining the RNA sequencing and RNA structure feature data of Wei with the speed and accuracy of Pearce’s DeepFoldRNA RNA folding simulations would create a predictable method using machine learning methods and secondary structure data and experimental chemical probing data (i.e., SHAPE-Seq and DMS-seq) for predicting tertiary RNA structure of an RNA molecule.
Claim(s) 6-7 are rejected under 35 U.S.C. 103 as being unpatentable over Pearce in view of Wei, as applied to claims 1-5, 8-10, 13-15, 20-25, 28-29, 33-34, and 36-43 above, and in further view of Marinus et al. (Biochemistry (Easton), 2019-06, Vol.58 (23), p.2655-2664).
Pearce and Wei teach claims 1-5, 8-10, 13-15, 20-25, 28-29, 33-34, and 36-43.
Pearce in view of Wei teach methods for predicting a tertiary structure of an RNA molecule of interest.
Pearce and Wei do not teach claims 6-7.
Marinus et al. (Marinus) teaches using SHAPE reagents (i.e., probes) for evaluating RNA structure [abstract]. Marinus teaches using NAI, I6, I5, 1M4, B5, 6A3, NIC and 2A3 as chemical probing agents.
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Pearce in view of Wei, and in further view of Marinus because Marinus teaches using chemical mapping agents NAI, I6, I5, 1M4, B5, 6A3, NIC and 2A3 for RNA analysis. One of ordinary skill in the art would recognize Marinus teaches methods for using SHAPE and DMS but Marinus teaches using the specific probing agents, 2A3, for RNA analysis for both in vivo and in vitro experimental conditions. Here, one of ordinary in the would be motivated to combine Pearce in view of Wei, and in further view of Marinus because Marinus teaches that the chemical agent, 2A3, will take over conventional SHAPE methods as 2A3 probing agent due to 2A3’s extreme efficiency and accuracy as compared to other probing agents (i.e., NAI, I6, I5, 1M4, B5, 6A3, NIC). Thus, there is a reasonable expectation of success to combine predicted structures of Pearce in view of Wei’s data sets, and in further view of Marinus’s SHAPE chemical probing agents because Marinus teaches that chemical probing agent 2A3 shows improvement with respect to other state of the art probing such as NAI which can be applied to the predicted structure of Pearce. Therefore, applying the SHAPE and DMS probing agents of Marinus to the predicted RNA structure Pearce using the datasets of Wei (i.e., experimental (i.e., SHAPE and DMS) and native data) would create an improved, efficient, and accurate predictable method steps for utilizing SHAPE and DMS chemical probing agents (i.e., NAI, I6, I5, 1M4, B5, 6A3, NIC and 2A3) for predicting RNA structures (i.e., tertiary, 3D).
Claim(s) 16-19, and 44 are rejected under 35 U.S.C. 103 as being unpatentable over Pearce in view of Wei, as applied to claims 1-5, 8-10, 13-15, 20-25, 28-29, 33-34, and 36-43 above, and in further view of Bliss et al. (RNA biology, 2020-09, Vol.17 (9), p.1324-1330).
Pearce and Wei teach claims 1-5, 8-10, 13-15, 20-25, 28-29, 33-34, and 36-43.
Pearce in view of Wei teach methods for predicting a tertiary structure of an RNA molecule of interest.
Pearce and Wei do not teach claim 16-19, and 44.
Bliss et al. (Bliss) teaches using neural network model for predicting SHAPE score (i.e., chemical mapping data) [Bliss, page1325 left col], as in claim 16.
Pearce
Pearce teaches models were produced using the assistance of experimental SHAPE data to elucidate secondary structure and contact information [Pearce page 6 Puzzle 6]. Pearce teaches the model from the RNA Puzzle 12 (i.e., tertiary RNA: ydaO riboswitch) was produced using fast-track experimental SHAPE data and multidimensional chemical mapping [Pearce page 6 Puzzle 12], as in claim 18.
Obvious claim(s): 17, 19, and 44
Pearce teaches models were produced using the assistance of experimental SHAPE data to elucidate secondary structure and contact information [Pearce page 6 Puzzle 6]. Pearce teaches model of RNA Puzzle (i.e., tertiary RNA: ydaO riboswitch) was produced using fast-track experimental SHAPE data and multidimensional chemical mapping [Pearce page 6 Puzzle 12]. Pearce teaches modeling both native RNA and its predicted structure using DeepFoldRNA [Pearce page 18 fig 3, page 20 fig 5]. Bliss teaches using a neural network model for predicting SHAPE score (i.e., chemical mapping data) [Bliss, page1325 left col], as in claim 17. Here, although Pearce and Bliss do not teach predicting the chemical mapping data for the RNA molecule of interest using the predicted tertiary structure of the RNA molecule of interest it would be obvious to analyze the native and predicted structures of Pearce together utilizing the SHAPE scoring system of Bliss so that the models can be tested against and refined to produce a further models.
Bliss teaches using neural network model for predicting SHAPE score (i.e., chemical mapping data) [Bliss, page1325 left col]. Pearce teaches modeling both native RNA and its predicted structure using DeepFoldRNA [Pearce page 18 fig 3, page 20 fig 5]. Pearce teaches using embeddings for predicting a final model [Pearce page 16 fig 6]. Pearce teaches the network architecture uses embeddings [Pearce page 8 network architecture], as in claim 19. Here, although Bliss and Pearce do not teach predicting the chemical mapping data for the RNA molecule of interest and the predicted tertiary structure it would be obvious to analyze the Puzzle and DeepFoldRNA models of Pearce using the deep learning model of Bliss to output SHAPE scores of the 3D/tertiary RNA structures (i.e., chemical mapping data).
Pearce teaches model of RNA Puzzle (i.e., tertiary RNA: ydaO riboswitch) was produced using fast-track experimental SHAPE data and multidimensional chemical mapping [Pearce page 6 Puzzle 12], as in claim 44 (a). Bliss teaches using 194 RNA’s for training and tests sets [Bliss, page 1325 table]. Wei also teaches using data sets from 18 experimental methods such as DMS-seq, SHAPE-Seq, SHAPE-MaP, and icSHAPE [Wei, page 6 right col RNA secondary structure], claim 44 (b). Here, although Wei, Bliss, and Pearce do not directly teach chemical probing experiments different from those used to originate the chemical mapping data for a first plurality of RNA molecules it is obvious because Wei, Bliss, and Pearce teach analyzing different RNA structures and the SHAPE and DMS data also contains RNA data for different RNA structures would have requires different chemical probing methods (i.e., SHAPE-Seq, DMS-seq).
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Pearce in view of Wei, and in further view of Bliss because Bliss teaches using deep learning methods for predicting RNA SHAPE scores. One of ordinary skill in the art would recognize that Bliss teaches methods for determining chemical mapping data scores that can help with RNA secondary structure predictions that inherently take in-vivo folding properties into account which can be utilized to target RNA molecules associated with specific disease pathways. Here, one of ordinary skill in the art would be motivated to combine the predicted structures of Pearce and the chemical mapping data (i.e., SHAPE-Seq/DMS-seq) of Wei with the pipeline of Bliss because Bliss teaches augmenting machine learning model with additional biological knowledge (i.e., chemical mapping and structural data of Wei and SHAPE score of Bliss ) should improve the prediction accuracy. Thus, combining the methods of Pearce in view of Wei, and in further view of Bliss would yield a predictable improved RNA tertiary structure analysis for predicting RNA tertiary structure using SHAPE scores (i.e., chemical mapping data) because Bliss teaches the combination of data would improve prediction accuracy .
Claim(s) 30-32 are rejected under 35 U.S.C. 103 as being unpatentable over Pearce in view of Wei, as applied to claims 1-5, 8-10, 13-15, 20-25, 28-29, 33-34, and 36-43 above, and in further view of Childs-Disney et al. (Nature Reviews Drug Discovery, 2022-10, Vol.21 (10), p.736-762.
Pearce and Wei teach claims 1-5, 8-10, 13-15, 20-25, 28-29, 33-34, and 36-43.
Pearce in view of Wei teach methods for predicting a tertiary structure of an RNA molecule of interest.
Pearce and Wei do not teach claims 30-32.
Childs-Disney et al. (Childs-Disney) teach sequence conservation of the 5′- end of beta coronaviruses suggested a functional role, and this region is predicted to fold into six stem–loop structures (SL1–SL6). Childs-Disney teaches one amiloride (i.e., drug) bound to SL6 (i.e., tertiary structure), as determined by a dye displacement assay and NMR spectral studies [Childs-Disney, page 748 right col], as in claim 30.
Childs-Disney teaches one amiloride (i.e., drug) bound to SL6 (i.e., tertiary structure), as determined by a dye displacement assay and NMR spectral studies [Childs-Disney, page 748 right col], as in claim 31. Here, a section/subsection of the six stem–loop structures (SL1–SL6), SL6, was targeted by the drug (i.e., amiloride).
Childs-Disney reviews methods for targeting RNA structures with small molecules [Childs-Disney abstract]. Childs-Disney teaches methods to identify or design small molecule RNA binders and a method to design chemical matter that binds to RNA [Childs-Disney pages 740-741 fig 2]. Childs-Disney teaches methods of validating a small molecule that targets RNA [Childs-Disney page 741-742 fig 3]. Childs-Disney teaches strategies for lead optimization of RNA-targeted small molecules [Childs-Disney page 747 fig 4]. Childs-Disney teaches tools for developing (i.e., formulating) small molecule therapeutics targeting RNA [Childs-Disney page 756], as in claim 32.
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Pearce in view of Wei, and in further view Childs-Disney because Childs-Disney reviews methods for targeting RNA structures with small molecules [Childs-Disney page 756]. One of ordinary skill in the art would recognize that Childs-Disney provides different methods, techniques, and examples to develop RNA molecules into drugs. Here, one of ordinary skill in the would be motived to choose one or more of the methods of Childs-Disney to develop/formulate a drug because Childs-Disney teaches general methods (i.e., fluorescence based assays, microarray-based assay [Childs-Disney page 741 fig 2]) that can be applied to the predicted structures of Pearce and Wei for identifying if a small molecule drug can target an RNA molecule. Thus, one of ordinary would expect a reasonable success using the methods of Childs-Disney to develop a drug to target an RNA molecule because Childs-Disney reviews methods known in the art for evaluating and developing/formulating predicted small molecules that can target said predicted RNA structures so that a diseases’ biochemical pathway could be stimulated or inhibited as a result of the small molecule. Therefore, combining the DeepFoldRNA program of Pearce trained with the combined data sets of Wei would create and output a predicted RNA molecule that can be subsequently utilized and analyzed for developing a small molecule drugs that can target said RNA molecule because Childs-Disney reviews known methods available for identifying/developing a small molecule can bind to said target RNA molecule.
Conclusion
Claims 1-10, 12-34 and 36-44 are rejected.
No claims are allowed.
Finality
This Office action is a Non-Final action. A shortened statutory period for reply to this action is set to expire THREE MONTHS from the mailing date of this action.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOSEPH C PULLIAM whose telephone number is (571)272-8696. The examiner can normally be reached 0730-1700 M-F.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Karlheinz Skowronek can be reached at (571) 272-9047. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/J.C.P./Examiner, Art Unit 1687
/Anna Skibinsky/
Primary Examiner, AU 1635
Prosecution Insights