This manuscript evaluates sure screening, a class of filter-based feature selection methods which provide analytical guarantees for true feature set retention and identified BcorSIS as the most effective and computationally efficient screening method across various omics datasets.
Abstract
In recent years, high dimensional omics analyses have become more commonplace for investigating complex biological systems. Typically, these studies attempt to identify key biomolecules associated with a particular biological process. Often, machine learning (ML) is used to identify these biomolecules, typically by learning which biomolecules are highly predictive of a treatment, biological outcome, or phenotype. A major challenge of applying ML to high throughput omics is overcoming noise when sample size is limited and unbalanced with respect to tens of thousands of biomolecules measured. Thus, feature selection (the process of reducing the number of predictors) is both a critical and common step in the ML analysis pipeline. While much attention has been given to embedding and wrapping techniques for feature selection in the omics space, filter-based methods for model-free feature selection have appealing theoretical properties. This manuscript evaluates sure screening, a class of filter-based feature selection methods which provide analytical guarantees for true feature set retention. Here, we cover existing feature screening methods based on the sure screening principal, available software, methods to improve feature screening, and contextualize feature screening in the larger discussion of feature selection for omics data analysis. Additionally, a suite of model-free sure screening approaches is applied and compared for several omics biomedical applications in a ML classification context. We identified BcorSIS as the most effective and computationally efficient screening method across various omics datasets, consistently outperforming others like CSIS and DCSIS in runtime.
Accurate prediction of drug sensitivity in cancer cell lines is vital for precision oncology and patient-specific therapies. However, many computational approaches fail to integrate multi-modal biological and chemical features and often struggle with high-dimensional, imbalanced pharmacogenomic data, limiting predictive accuracy and interpretability. To address these challenges, we developed a machine learning framework that integrates pharmacogenomic profiles-including mutation status, copy number alterations, and microsatellite instabil-ity-with molecular fingerprints and descriptors of 85 anticancer drugs, generated using PaDEL from SMILES strings. Data from 40 breast cancer cell lines in the Genomics of Drug Sensitivity in Cancer (GDSC) dataset were employed. A threestage feature selection strategy combining Boruta, mRMR, and XGBoost was applied to reduce drug feature dimensionality while retaining 130 cell line features. Multiple models were trained, and LightGBM, optimized with grid search, class weighting, and 3-fold cross-validation, demonstrated superior performance in handling severe class imbalance (233 sensitive vs. 3167 resistant samples). LightGBM achieved training AUROC $=0.9455$, AUPRC $\boldsymbol{=} \mathbf{0. 5 1 4 8}$, Accuracy $\boldsymbol{=} \mathbf{0. 8 4 1 5}$, F1-score = 0.4481, Recall = 0.9409, and MCC = 0.4732, underscoring its suitability for sparse biomedical datasets. Model interpretation with SHapley Additive exPlanations (SHAP) highlighted BRCA-related features, identifying cnaBRCA25 (not mutated) as a resistance marker and cnaBRCA47 (mutated) as a context-dependent biomarker, consistent with their roles in DNA repair pathways. Overall, this framework demonstrates the value of multi-modal integration and interpretable machine learning in pharmacogenomics. While results are promising, validation on larger and independent cohorts is essential to establish clinical relevance.
D. Kumari, Aiman, Sakshi Singh et al.· Annual International Compute...· 0 citations
Logistic regression remains a widely used classification method due to its interpretability and computational efficiency, but its direct application to high-dimensional biomedical data is limited when the number of features greatly exceeds the number of samples. In this paper, we propose a reformulated logistic regression framework designed for feature selection and classification in complex high-dimensional settings. The method is evaluated on three biomedical datasets, including scenarios with tens of thousands of attributes and substantially fewer samples. Across these datasets, the proposed approach achieved clear separation between control and disease groups while selecting a compact set of features. Several selected features were consistent with previously reported disease-associated markers, supporting the biological plausibility of the model, while additional selected features suggest potential novel candidates for further investigation. These results indicate that the proposed framework may provide an interpretable and computationally efficient alternative for feature selection in high-dimensional computational biology applications.
Rossana O. Souza, W. Rodrigues, Bráulio Couto et al.· bioRxiv· 0 citations
MB-SupCon-cont improves prediction accuracy by incorporating a generalized contrastive loss function that defines similarity and dissimilarity for continuous responses using three distance-based weighting methods, and provides superior representation learning and improves data visualization in lower-dimensional spaces.
Sen Yang, Shidan Wang, Yiqing Wang et al.· Frontiers in microbiomes· 0 citations
Bulk RNA-seq data suffers from the issues of “high dimensionality and small sample size,” which limits its application in disease research. This paper proposes a dynamic data augmentation method based on Layer-wise Relevance Propagation (LRP) aimed at improving classification performance and biological interpretability under small-sample conditions. The method utilizes the LRP algorithm to calculate the contribution weight of each gene to the classification result and uses this weight to guide sample generation. By systematically amplifying biologically meaningful signals, it constructs semantically reliable augmented samples, avoiding the semantic distortion caused by traditional random perturbations. Simultaneously, a dynamic augmentation mechanism is introduced that tightly couples sample generation with model training, providing difficult-to-classify samples with multiple iterative optimization opportunities and forming a virtuous cycle where classification performance and augmentation quality improve synergistically. On this basis, population-level gene biomarkers are identified from the trained model. Innovatively, an open-environment enrichment analysis method is proposed—that is, instead of being limited to a few feature genes of a single subtype, the union of feature genes from all disease subtypes is taken for enrichment analysis, revealing shared biological pathways from a systems-level perspective and providing a more comprehensive interpretation for subtype-specific mechanism research. Experimental results show that this method effectively improves classification accuracy, and through this open-environment enrichment approach, six hub genes were identified as gene markers for IgAN.
Fang Zheng, Juanjuan Zhao, Baoping Jia et al.· Journal of Biological Engine...· 0 citations
Survival analysis is a fundamental technique in biomedical research for modeling time-to-event data. It enables the identification of prognostic factors in disease, compares survival outcomes across treatment groups, and performs targeted treatment selection. A variety of machine learning (ML) approaches to survival analysis have emerged to complement classical statistical methods, especially for high-dimensional datasets with complex, nonlinear interactions between features. However, using survival ML methods requires addressing challenges such as censoring-unaware evaluation, overfitting, selecting performance metrics, and data leakage. To address these and other difficulties in using survival ML models, we developed the mlsurv software package. mlsurv is an open-source Python package built around three major design principles: 1) methodological rigor, including evidence-based model selection, leakage-free pipelines, and multi-metric evaluation, 2) multi-scale evaluation and interpretation, including population and subpopulation evaluation, patient-level explanations, and feature analysis, and 3) automated trust and transparency, including limitation flagging and TRIPOD+AI-aligned reporting. mlsurv bundles ten models spanning linear, ensemble, kernel, and deep learning families within a unified software package. We demonstrate mlsurv on the Chowell immunotherapy cohort (n=1,479). The survival-trained models achieve a test concordance index of 0.73 for overall survival prediction. Further, risk scores strongly correlate with the response-trained LORIS clinical score (|{rho}| up to 0.84), reflecting the overlap between prognostic and predictive signal. mlsurv enables biomedical researchers to conduct rigorous, multi-model survival analysis and benchmarking using minimal code with default best practices rather than implementing custom scripts and methodological safeguards from scratch.
A. Pybus, J. Qiu, P. C. Morais Lyra et al.· medRxiv· 0 citations
This PhD thesis investigates the reliable application of machine learning (ML) and deep learning (DL) methods to omics data, with particular emphasis on high-dimensional, low-sample-size settings commonly encountered in foodomics and agronomy. The work focuses on improving methodological rigor, interpretability, and reproducibility in applied omics research. The dissertation is organized into two main parts. The first part addresses methodological aspects of ML for omics data, including evaluation strategies and hybrid modeling approaches. In particular, it examines the interaction between cross-validation and early stopping in neural network training, identifying common pitfalls such as information leakage and biased performance estimation. Furthermore, the thesis explores hybrid neural network architectures that integrate mechanistic domain knowledge into data-driven learning, framing the problem as a multi-objective optimization task that balances predictive accuracy with mechanistic consistency. The second part focuses on applied case studies in foodomics and agronomy. It presents robust and explainable deep learning models for SNP-based phenotype prediction, demonstrating statistically significant improvements in predictive performance through adaptive optimization, regularization, and data augmentation strategies. The work also employs SHAP-based explainability methods to identify biologically relevant features and ensure transparent interpretation of model predictions. In addition, the thesis introduces a research-stage MLOps framework for organizing complex omics machine learning workflows, improving experiment traceability, reproducibility, and comparability. Overall, the dissertation contributes to the development of more reliable, interpretable, and reproducible ML methodologies for omics data analysis by combining methodological innovation with practical applications in foodomics and agronomy.