Unlocking patterns in limited omics data

Le résumé fourni par la source

Accurate clinical diagnostic tests are essential for effective cancer treatment. Machine learning (ML) models applied to omics data - such as genomic, transcriptomic, or proteomic profiles - hold promise for enhancing diagnostic practices. However, training reliable ML models is often challenged by limitations in available datasets. Solid tissue biopsy remains the gold standard for tumor sampling, but liquid biopsy is being applied more frequently due to its non-invasive, safer, and easier procedure. Yet, liquid biopsy data presents unique challenges. The sparsity of mutation profiles in liquid biopsy data can hinder model training as the underlying patterns may be determined less reliably and can differ from those in solid tissue biopsy. Domain shift represents another challenge, where datasets from different sources have varying distributions due to differences in experimental protocols or patient populations. This shift can reduce model performance when applied to new, unseen data. Additionally, small sample sizes in clinical cohorts often lead to high-dimensional datasets (e.g., gene expression data), causing the “curse of dimensionality” and overfitting. Class imbalance, where one class dominates the training set, can also decrease model performance on (the) minority class(es). This thesis explores ML applications for cancer omics datasets affected by sparsity, domain shift, small sample sizes, and class imbalance. In Chapter 2, a deep learning model, originally trained on solid tissue biopsy data, is adapted to predict the “cell-of-origin” of cancer using liquid biopsy data. Synthetic datasets model the sparse mutation profiles of liquid biopsies, and data augmentation combined with diverse feature types are integrated to improve model performance. The adapted deep learning model achieves classification accuracy on synthetic sparse data comparable to its performance on solid tissue data, demonstrating the potential of liquid biopsies for advanced cancer diagnostics. In Chapter 3, the same classification model is evaluated on a mixed primary-metastatic solid biopsy dataset. Synthetic oversampling techniques are tested to mitigate class imbalance, while subspace-centric and data-centric domain adaptation methods address domain shift. The findings offer strategies to enhance cancer type classification models under these conditions. In Chapter 4, a novel ML approach predicts treatment response to abiraterone and enzalutamide (ARSI therapies) in metastatic castration-resistant prostate cancer (mCRPC) patients. Using whole genome and transcriptome sequencing data from a small clinical cohort (n=155 and n=113, respectively), potential predictive features are identified. Models combining prior treatment information with genomic markers or transcriptomic data projected in a lower dimensional space show promise. With further validation, this method could guide treatment decisions and identify patients most likely to benefit from new therapies. In Chapter 5, the broader implications of this thesis are discussed, including limitations and future directions for ML in cancer diagnostics, emphasizing the need for robust, generalizable models to improve patient outcome. This work highlights the potential of ML to overcome key challenges in cancer omics data, from sparsity and domain shift to small sample sizes and class imbalance. By leveraging synthetic data, domain adaptation, and feature selection, ML models can be optimized for clinical applications, ultimately advancing personalized cancer diagnostics and treatment.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.