News

Distribution-based feature extraction from copy number and structural variants in whole-genome data

Date
August 5, 2022
No items found.

With entire genomes at hand, scientists can pinpoint DNA fragments responsible for different cancers and predict patient responses to cancer treatments. However, the sheer volume of whole-genome data makes it difficult to encode the characteristics of genomic variants as features for machine learning algorithms.

That is why, at MNM, we are working on different ways to tackle this problem. In our recent study, we propose three feature extraction methods, called DBFE, that facilitate classifier learning from distributions of genomic variants. The proposed approaches use binning, clustering, and kernel density estimation to produce features that discriminate between two groups of patients. Experiments on genomes of ovarian, lung, and breast cancer patients show that the proposed approaches are capable of automatically identifying genomic biomarkers associated with cancer subtypes and clinical response to oncological treatment. Moreover, the extracted features are highly interpretable and can be used alongside unsupervised learning methods to analyze genomic samples.

Distribution-based feature extraction from copy number and structural variants in whole-genome data
August 9, 2022

You may also like

Europe Biobank Week 2021

Almost a week has passed since the Europe Biobank Week 2021, and yet we are still impressed! We have attended great lectures, in which the speakers approached the topic of biobanking from different perspectives.

Distribution-based feature extraction from copy number and structural variants in whole-genome data

With entire genomes at hand, scientists can pinpoint DNA fragments responsible for different cancers and predict patient responses to cancer treatments. However, the sheer volume of whole-genome data makes it difficult to encode the characteristics of genomic variants as features for machine learning algorithms.

Greetings from ECML PKDD 2022!

Maciej Piernik presented one of our core algorithms at the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases.

MNM Bioscience eyes cancer drug target discovery based on its AI algorithms

MNM Bioscience began as a bioinformatics company that curates data and then builds artificial intelligence-based algorithms, but it is now looking to branch out into oncology drug target discovery.