Clonal Neoantigen Selection Using Bayesian Clonality Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately determine whether tumor-specific mutations are clonal, leading to limited therapeutic efficacy and potential relapse or metastasis, as they often target subclonal mutations, and there is a need for methods to identify clonal neoantigens for effective cancer therapy and prognosis.
Innovation Solution
A method using a Bayesian framework to classify tumor-specific mutations as clonal based on sequence data, incorporating somatic copy number aberration data and accounting for uncertainty in copy number calls, providing a posterior probability of clonality through numerical integration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing methods are used to determine clonality of tumor-specific mutations, then the analysis can be performed, but the accuracy is insufficient leading to limited therapeutic efficacy
Solution Approach 1:
The method changes the parameters used for clonality assessment by incorporating somatic copy number aberration data and using a Bayesian framework that calculates posterior probabilities. This transforms the analysis from simple mutation frequency counting to a multi-parameter statistical model that accounts for copy number variations, sequencing depth, and tumor purity, thereby improving accuracy and reliability of clonality determination
Solution Approach 2:
The patent introduces an intermediary statistical framework (Bayesian analysis) that mediates between the raw sequence data and the final clonality classification. This intermediary layer processes multiple inputs (read counts, copy number data, tumor fraction) through likelihood calculations and prior probabilities to produce posterior probability assignments, enabling more accurate and reliable clonality assessment
2Productivity
If subclonal mutations are targeted, then some tumor cells are treated, but relapse or metastasis occurs because unaffected clones remain
Solution Approach 1:
The method extracts and identifies only the clonal mutations (those present in all tumor cells) from the total set of tumor-specific mutations. By separating clonal from subclonal mutations through Bayesian analysis of sequence data and copy number aberrations, the system enables targeted therapies to focus exclusively on mutations that affect all tumor cells, thereby preventing relapse and metastasis while maintaining therapeutic productivity
3Measurement precision
If a rigorous statistical framework is used to classify mutations, then accuracy improves, but computational complexity increases
Solution Approach 1:
The patent segments the computational framework into distinct modular components: (1) input data processing (sequence data, copy number data, tumor fraction), (2) likelihood calculation for different clonality scenarios, (3) Bayesian posterior probability computation, and (4) clonality classification. This segmentation allows each component to be optimized independently and facilitates implementation while maintaining high accuracy through rigorous statistical analysis
Data Source
AI summary
A method of treating a subject having been diagnosed as having cancer with an immunotherapy is described. The method comprises targeting one or more clonal neoantigens selected using a method comprising determining whether a tumour-specific mutation is likely to be clonal in a subject. The method comprises providing sequence data from one or more samples from the subject comprising tumour genetic material, the sequence data comprising for each of the one or more samples, and determining the likelihood that the tumour-specific mutation is clonal as a posterior probability depending on: a prior probability of the mutation being clonal, and the probabilities of observing the sequence data if the tumour-specific mutation is (i) clonal and (ii) non-clonal, in view of a tumour fraction for each of the one or more samples and one or more candidate joint genotypes.