Prosecution Insights
Last updated: April 19, 2026
Application No. 17/986,559

APPLYING A LAYERED APPROACH TO DETERMINING MOLECULAR RETROSYNTHETIC ROUTE USING A NEURAL NETWORK

Non-Final OA §112
Filed
Nov 14, 2022
Examiner
HYUN, PAUL SANG HWA
Art Unit
1796
Tech Center
1700 — Chemical & Materials Engineering
Assignee
Tencent Technology (Shenzhen) Company Limited
OA Round
2 (Non-Final)
70%
Grant Probability
Favorable
2-3
OA Rounds
3y 5m
To Grant
99%
With Interview

Examiner Intelligence

Grants 70% — above average
70%
Career Allow Rate
582 granted / 834 resolved
+4.8% vs TC avg
Strong +36% interview lift
Without
With
+36.0%
Interview Lift
resolved cases with interview
Typical timeline
3y 5m
Avg Prosecution
38 currently pending
Career history
872
Total Applications
across all art units

Statute-Specific Performance

§101
1.0%
-39.0% vs TC avg
§103
38.5%
-1.5% vs TC avg
§102
23.0%
-17.0% vs TC avg
§112
31.0%
-9.0% vs TC avg
Black line = Tech Center average estimate • Based on career data from 834 resolved cases

Office Action

§112
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 . Response to Amendment The amendment filed on December 29, 2025 is acknowledged. Claims 1-5, 7-16 and 18-20 remain pending wherein claims 10 and 11 remain withdrawn. Applicant amended claims 1, 4, 7, 12 and 18. Applicant also filed a new abstract. Response to Arguments Applicant’s arguments with respect to the patentability of the claims have been considered but they are moot. The Office action includes the new grounds of rejection unrelated to the issues specifically challenged in the arguments. Claim Objections Claim 4 is objected to because of the following informalities: In the penultimate line of claim 4, the limitation “functions” should be changed to “function”. The claim previously recites “function of the respective first molecule”, meaning there is a one-to-one correspondence between a function and a molecule. Claim 4 recites “molecular cost value reference information” which is nominally inconsistent with “cost value reference information” recited in claim 2. Nomenclature should be consistent throughout the claims to obviate potential indefiniteness. Appropriate corrections are required. Claim Interpretation The following limitations, unless explicitly defined in the claims, are being interpreted based on their definitions provided in the specification: “Molecular expression information”: Information representing a three-dimensional chemical structure of a molecule. “Disassembly path”: A path that requires the least cost to dissemble a molecule until a target disassembly condition is met. “Target disassembly condition”: A path depth of a disassembly path obtained for each molecule based on at least one disassembly level that is less than a target path depth and there is no disassembly method for the obtained molecules, or a path depth of the disassembly path obtained for each molecule based on at least one disassembly level that is equal to the target path depth. “Disassembly task”: A process of gradually dissembling a molecule into at least one simpler and easier-to-synthesize molecule according to a disassembly path. “cost value reference information”: A synthetic accessibility score, which is a scoring system for quickly assessing the ease of synthesis of a compound based on the complexity and ubiquity of the molecules of the compound. Claim Rejections - 35 USC § 112 The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action. Claims 1-5, 7-9, 12-16 and 18-20 are rejected under 35 U.S.C. 112(b) as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention. Claims 1 and 12 recite a step of training without specifying what is being trained. In addition, the nexus between the “training” and the “generating” steps at the end of the claims is unclear. According to Applicant’s remarks, the amendment to the end of the claims was made to overcome the outstanding 35 U.S.C. 101 rejection, yet the claims do not specify a nexus between the abstract idea of “training” and the “generating” steps to unequivocally integrate the abstract idea into a practical application. In other words, without the nexus, it is unclear whether the abstract idea of “training” is integrated into a practical application so as to overcome the outstanding 35 U.S.C. 110 rejection. Currently, the claims recite a step of training, and transition to unrelated steps of generating the target neural network and the retrosynthetic route. Claims 2 and 13 recites, “in response to a determination that a disassembly level of each of the first molecules is complete…”. There is no context for the recitation. The claims do not previously recite a step of dissembling the first molecules, let alone determining whether the disassembly level is complete, and it is also unclear who or what is performing the determination. Likewise, the recitation “in response to a determination that a disassembly task…satisfies a disassembly condition” lacks context. The claims do not previously recite a step of determining whether a disassembly task satisfies a disassembly condition. Claim 8 recites a step of “training a first neural network”. Given that the first neural network is first introduced in claim 8 and it is being trained, it is unclear what is performing the steps of the claim (i.e. what is training the first neural network?). This rejection is related to the issue identified in claim 1 above regarding the limitation “performing training” (i.e. what is being trained and by what?). Relatedly, claim 12 recites a processing circuitry that performs the steps recited in the claim and the dependent claims. The claim subsequently recites that the circuitry performs training and it also recites a target neural network, and the dependent claims recite a first neural network and a second neural network trained by the circuitry. The relationship between the circuitry of claim 12 and the networks recited in the claims is unclear. Presumably each network comprises its own processing circuitry. That said, it is unclear to what the processing circuitry of claim 12 belongs, and how it trains/generates the networks recited in the claims. Claims 9 and 20 are indefinite because they recite circular logic. The claims recite a step of obtaining predicted cost value information of each third molecule by using the second cost dictionary. Yet, according to claims 1 and 12, the second cost dictionary is obtained by using cost value information of each third molecule. In other words, claims 9 and 20 recite a step of obtaining predicted cost value information by using cost value information. It is unclear why or how information is predicted if said information is known. Claims not explicitly rejected are rejected due to dependency. Allowable Subject Matter While no claim is identified as being allowable due to the rejections set forth above, no prior art is applied against the claims. The closest art remains Wlodarczyk-Pruszynski et al. (“W-P”) (US 2021/0125691 A1). W-P discloses a method of training a neural network configured to determine a molecular retrosynthetic route (see abstract and [0139]), the method comprising (see [0140] and Fig. 11): determining a first disassembly path (reaction) for each of a plurality of first molecules (step 1102 in Fig. 11) based on molecular expression information (structure or probability of reaction) of the respective one of the plurality of first molecules (see [0048]); obtaining and subsequently using a first cost dictionary (“selecting the most promising path/reaction”) based on the first disassembly paths of the first molecules (see [0139]-[0140]), the first cost dictionary comprising the molecular expression information and cost value information corresponding to each of the first molecules (see [0056] and [0081]), wherein the cost value information represents a cost required to disassemble a first molecule according to the first disassembly path of said first molecule (see [0077]-[0081]); determining molecular expression information of second molecules based on the first disassembly paths of the first molecules (step 1106), each of the second molecules being a molecule that is obtained by disassembling a corresponding first molecule based on the first disassembly path, wherein each of the second molecules is capable of being further disassembled (see step 1108); obtaining and subsequently using a second cost dictionary based on second disassembly paths of the second molecules (step 1108), the second cost dictionary comprising the molecular expression information and cost value information corresponding to each of the second molecules (see above), wherein the cost value information of each second molecule represents a cost required to disassemble the respective second molecule according to the second disassembly path (see above); and using the first and second cost dictionaries to train the neural network to generate a retrosynthetic route for a target molecule (step 1110), and to output cost value information corresponding to the retrosynthetic route (see abstract disclosing that synthesis pathways of the target molecule are ranked according to a cost estimation and the ranking is provided to a user). The method differs from the claimed invention in that W-P does not utilize Tanimoto coefficients to group or cluster molecules into classes, determine the cluster center of each class, and use the cluster center to determine a molecule representative of each class, as recited in claims 1 and 12. While W-P discloses the use of Tanimoto coefficients (i.e. Jaccard index) (see [0121]), it is not used for the purpose recited in the claims. Based on the disclosure of W-P and other relevant prior art, there is no motivation to modify the method to arrive at the claimed inventions. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to PAUL S HYUN whose telephone number is (571)272-8559. The examiner can normally be reached M-F 8:30-5:00. 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, Luan Van can be reached at 571-272-8521. 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. /PAUL S HYUN/Primary Examiner, Art Unit 1796
Read full office action

Prosecution Timeline

Nov 14, 2022
Application Filed
Sep 24, 2025
Non-Final Rejection — §112
Oct 30, 2025
Applicant Interview (Telephonic)
Oct 30, 2025
Examiner Interview Summary
Dec 29, 2025
Response Filed
Mar 10, 2026
Non-Final Rejection — §112 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12599127
VITRIFICATION DEVICE AND METHOD FOR PREPARING SAMPLE
2y 5m to grant Granted Apr 14, 2026
Patent 12599906
CONTAINER FOR SMALL LIQUID VOLUMES
2y 5m to grant Granted Apr 14, 2026
Patent 12590508
MANIPULATION OF FLUIDS, FLUID COMPONENTS AND REACTIONS IN MICROFLUIDIC SYSTEMS
2y 5m to grant Granted Mar 31, 2026
Patent 12558684
ELECTRONIC DEVICE
2y 5m to grant Granted Feb 24, 2026
Patent 12551888
DEVICES AND METHODS FOR EXTRACTION, SEPARATION AND THERMOCYCLING
2y 5m to grant Granted Feb 17, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

AI Strategy Recommendation

Get an AI-powered prosecution strategy using examiner precedents, rejection analysis, and claim mapping.
Powered by AI — typically takes 5-10 seconds

Prosecution Projections

2-3
Expected OA Rounds
70%
Grant Probability
99%
With Interview (+36.0%)
3y 5m
Median Time to Grant
Moderate
PTA Risk
Based on 834 resolved cases by this examiner. Grant probability derived from career allow rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month