公开(公告)号：US12009060B2

公开(公告)日：2024-06-11

申请号：US16362236

申请日：2019-03-22

Applicant: Merck Sharp & Dohme LLC ,  MSD Czech Republic s.r.o.

Inventor： Geoffrey D. Hannigan ,  David Prihoda ,  Jindrich Soukup ,  Christopher Harron Woelk ,  Danny A. Bitton

IPC: G16B30/00 ,  G06N3/044 ,  G06N3/045

CPC classification number: G16B30/00 ,  G06N3/044 ,  G06N3/045

Abstract: A BGC prediction system identifies candidate biosynthetic gene clusters (BGCs) within genomes using machine-learned models, such as a shallow neural network and recurrent neural network (RNN). A set of domains within a genome sequence are identified, each domain corresponds to a set of domain identifiers. A shallow neural network block is applied to each set of domain identifiers to produce a set of vectors. An RNN block is applied to the set of vectors to produce a BGC class score for each domain. The RNN block was trained using an identified set of positive vectors, which represents known BGCs, and a synthesized set of negative vectors, which is unlikely to represent BGCs. Candidate BGCs are selected by averaging BGC class scores across genes within a domain and comparing the average BGC class scores to a threshold. The candidate BGCs are provided for display on a user interface.

公开(公告)号：US20240105285A1

公开(公告)日：2024-03-28

申请号：US17954302

申请日：2022-09-27

Applicant: Merck Sharp & Dohme LLC ,  MSD Czech Republic s.r.o.

Inventor： Pavel Vácha ,  Richard Dobis ,  Ondrej Micka ,  Nikita Vostrosablin ,  Martin Spale ,  Josef Navrátil ,  Jaroslav Cerman ,  Brian Lahue ,  Thomas James Struble ,  Zhengwei Peng ,  John Steele Debenham ,  Danny Asher Bitton ,  Petr Mejzlik

IPC: G16C20/10

CPC classification number: G16C20/10

Abstract: A reaction pathfinder system can leverage historical chemical reaction data and determine a synthesis route in a reaction network. The reaction pathfinder system can account for various performance criteria of chemical reactions such as a highest yield for a molecule or a minimal number of reaction steps. A reaction pathfinder system receives a user request for a synthesis route including one of a user-specified molecule or a user-specified reaction. The reaction pathfinder system may then query a reaction network that has various synthesis routes and represents reactions of reactants to produce respective molecules. The reaction network may be composed of molecule nodes and reaction nodes. The reaction pathfinder system determines, using the reaction network, the synthesis route from the synthesis routes to fulfill the user's request.

公开(公告)号：US20250095778A1

公开(公告)日：2025-03-20

申请号：US18886133

申请日：2024-09-16

Applicant: Merck Sharp & Dohme LLC ,  MSD Czech Republic s.r.o. ,  MSD Biotech B.V.

Inventor： Paul Alan Duncan ,  Julia Meredith Maritz ,  Geoffrey Daniel Hannigan ,  Christopher Harron Woelk ,  Christopher James Wang ,  Jack Benjamin Baker ,  Vanessa Vazquez Sarathy ,  Ron Šmeral ,  Ales Vondra ,  Ondrej Klempir ,  Ondrej Tupa ,  Anna Gromek ,  Danny Asher Bitton ,  Jakob Moritz Goldmann

IPC: G16B30/00

Abstract: A sequencing data analysis platform can process datasets that include a large number of sequence read. The reads are aligned to one or more reference genomes. Due to sequencing errors, sequencing noise, or genuine differences between a reference genome and the individual species being sequenced, this mapping process may tolerate a certain number of mismatches, insertions, or deletions. The sequencing data analysis platform provides a set of tools for analyzing and visualizing the sequence reads.

公开(公告)号：US20240095636A1

公开(公告)日：2024-03-21

申请号：US18465645

申请日：2023-09-12

Applicant: MSD Czech Republic s.r.o. ,  Merck Sharp & Dohme LLC

Inventor： Ioannis Spyroglou ,  Chad Rutt

IPC: G06Q10/0635

CPC classification number: G06Q10/0635

Abstract: A clinical investigation management system monitors clinical investigations performed across departments and clinical investigators. The system employs a method for predicting timeliness in completion of clinical investigations. The method includes monitoring data of a clinical investigation performed by a clinical investigator. The method includes applying a timeliness model to the data to determine a timeliness prediction of the clinical investigation. The method includes identifying one or more interventive actions based on the timeliness prediction. The method includes generating a notification including the timeliness prediction and the identified one or more interventive actions. The method includes transmitting the notification to a client device of a supervisor.

公开(公告)号：US11605025B2

公开(公告)日：2023-03-14

申请号：US16874232

申请日：2020-05-14

Applicant: MSD International GmbH ,  MSD Czech Republic s.r.o.

Inventor： Yingqi Peh ,  Kah Hin Chin ,  Shao Ying Choo ,  Sucitro Dwijayana Sidharta ,  Richard Dobis

IPC: G06N20/00 ,  G06F17/18

Abstract: As a data science project goes into the production stage, model maintenance to maintain model quality and predictive accuracy becomes a concern. Manual model maintenance by data scientists can become a time- and labor-intensive process, especially for large scale data science projects. An early warning system addresses this by performing systematic statistical and algorithmic checks for prediction accuracy, stability, and model assumption validity. A diagnostic report is generated that helps data scientists to assess the health of the model and identify sources of error as needed. Well-performing models can be automatically deployed without further human intervention while poor performing models trigger a warning or alert to the data scientists for further investigation and may be removed from production until the performance issues are addressed.

公开(公告)号：US20240290429A1

公开(公告)日：2024-08-29

申请号：US18654581

申请日：2024-05-03

Applicant: Merck Sharp & Dohme LLC ,  MSD Czech Republic s.r.o.

Inventor： Geoffrey D. Hannigan ,  David Prihoda ,  Jindrich Soukup ,  Christopher Harron Woelk ,  Danny A. Bitton

IPC: G16B30/00 ,  G06N3/044 ,  G06N3/045

CPC classification number: G16B30/00 ,  G06N3/044 ,  G06N3/045

Abstract: A biosynthetic gene cluster (BGC) prediction system identifies candidate BGCs within genomes using an iteratively trained machine-learned model. The system identifies, in a genome sequence, a set of domains, each identified domain corresponding to a set of domain identifiers. The set of domain identifiers corresponds to a set of vectors. The iteratively trained model is applied to the set of vectors to produce a BGC class score for each domain. The system selects candidate BGCs by averaging GBC class scores across genes within a domain and comparing the average BGC class scores to a threshold. The system predicts a molecular activity of biosynthetic products derived from the selected BGCs, and provides for display, on a user interface, the candidate BGCs and predicted molecular activity.