Pan-cancer platinum response predictor
Inventors
Abraham, Jim • Korn, Wolfgang Michael • Spetzler, David
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11842805-B2
Publication Date
2023-12-12
Expiration Date
Abstract
Comprehensive molecular profiling provides a wealth of data concerning the molecular status of patient samples. Such data can be compared to patient response to treatments to identify biomarker signatures that predict response or non-response to such treatments. This approach has been applied to identify biomarker signatures that correlate with response of cancer patients to platinum-based chemotherapy. Described herein are data structures, data processing, and machine learning models to predict a probability of benefit of a treatment for a disease or disorder of a subject having a particular set of biomarkers, as well as an exemplary application of such a model to precision medicine, e.g., to methods for selecting a treatment based on a molecular profile, e.g., a treatment comprising platinum therapy.
Core Innovation
The invention relates to a system and method for selecting a treatment for a cancer in a first subject by using copy numbers derived from output data generated by a next generation sequencer. The next generation sequencer sequences a biological sample comprising cancer cells from the first subject, and the obtained plurality of copy numbers includes a copy number for each of multiple groups of genes or proximate genomic regions thereto, including Group 1 through Group 5.
The obtained copy numbers are provided as input data to a predictive model that includes multiple machine learning models. Each machine learning model processes the obtained copy numbers for one of Group 1, Group 2, Group 3, Group 4, and Group 5, and generates group-specific data indicating whether the first subject is likely to benefit from a treatment that includes platinum therapy.
The system and method determine, based on the generated data from the multiple machine learning models, whether a majority indicates that the first subject is likely to benefit from the treatment that includes platinum therapy. Based on a determination that a majority indicates likelihood of benefit, the computers identify data that identifies the treatment that includes platinum therapy and provide output that identifies the platinum therapy; in related implementations, if a majority does not indicate likely benefit, an alternative treatment is identified and output, with the alternative constrained to exclude platinum therapy.
Claims Coverage
The document provides three independent claims: system, method, and non-transitory computer-readable storage media. Across these independent claims, the inventive features include obtaining copy numbers for five predefined gene groups, processing each group with a dedicated machine learning model within a predictive model, and selecting platinum therapy when a majority indicates likely benefit.
Multi-model platinum-benefit selection using gene-group copy numbers
A system comprising one or more computers and one or more memory devices storing instructions that cause the computers to obtain a plurality of copy numbers based on output data generated by a next generation sequencer from sequencing a biological sample comprising cancer cells from the first subject, and provide input data including the copy numbers to a predictive model including multiple machine learning models configured to process copy numbers for Group 1, Group 2, Group 3, Group 4, and Group 5.
Group-specific model processing to generate likelihood-of-benefit data
Processing the input data that includes the obtained copy numbers for each group through a respective machine learning model to generate corresponding data indicating whether the first subject is likely to benefit from a treatment that includes platinum therapy.
Majority determination to identify and output platinum therapy
Determining, based on the generated data, whether a majority of the multiple machine learning models indicates that the first subject is likely to benefit from the treatment that includes platinum therapy; based on that determination, identifying data that identifies the treatment that includes platinum therapy; and providing output that identifies the treatment that includes platinum therapy.
Computer-implemented method for selecting platinum therapy using predefined gene groups
A method for selecting a treatment for a cancer in a first subject comprising obtaining a plurality of copy numbers based on next generation sequencer output from sequencing a biological sample comprising cancer cells from the first subject, providing input data including the copy numbers to a predictive model including multiple machine learning models configured to process copy numbers for one of Group 1 through Group 5, processing the copy numbers for each group through a respective machine learning model to generate data indicating whether the first subject is likely to benefit from platinum therapy, and determining whether a majority indicates likely benefit to identify and output the treatment that includes platinum therapy.
Non-transitory media executing the platinum-benefit selection operations
One or more non-transitory computer-readable storage media storing instructions that, when executed by one or more computers, cause operations for selecting a treatment for a cancer in a first subject comprising obtaining plurality of copy numbers from next generation sequencer output, providing copy-number input to a predictive model including multiple machine learning models configured to process copy numbers for Group 1 through Group 5, processing each group through a respective machine learning model to generate data indicating likely benefit from platinum therapy, determining whether a majority indicates likely benefit, identifying data that identifies the treatment that includes platinum therapy, and providing output that identifies the treatment that includes platinum therapy.
All independent claims cover computer-implemented selection of a platinum-therapy treatment by converting next generation sequencer output into copy numbers for five predefined gene groups, processing each group’s copy numbers with a corresponding machine learning model within a predictive model, and selecting platinum therapy when a majority of model outputs indicates likely benefit.
Stated Advantages
Documented Applications
Selecting a treatment for a cancer in a first subject by identifying and outputting a treatment that includes platinum therapy when a majority of machine learning models indicates likely benefit.
Selecting a treatment when a majority of the machine learning models does not indicate likely benefit from a treatment that includes platinum therapy, including identifying and outputting an alternative treatment.
Interested in licensing this patent?