omics-ml-classifier

To train a machine learning classifier on gene expression data, provide a numeric feature matrix and a label file to train models like random_forest, logistic, svm, or xgboost, yielding cross-validated metrics and feature importances for biomarker ranking.

Yes, cross-validation evaluates multi-class classification models for omics data. The workflow supports two-class and multi-class labels, outputting cross-validated performance metrics, a held-out test set, and publication-ready ROC curves.

Predicting cancer subtypes from genomics feature matrices involves training supervised classifiers like svm or xgboost on the numeric expression data. The workflow preprocesses inputs, trains models, and outputs performance metrics with feature importance rankings.

Yes, the omics classifier workflow supports xgboost and random forest models, alongside logistic regression and svm. It trains these models using scikit-learn and xgboost dependencies, providing cross-validated metrics and feature importance outputs.

Generate publication-ready ROC curves from multi-omics classification results by training supervised classifiers on your numeric feature matrix. The skill outputs these visualizations alongside cross-validated metrics and a held-out test set using matplotlib and seaborn.

Omics classifier training requires a separate label file and numeric feature matrix because the supervised machine learning workflow needs distinct phenotype labels to map against expression features for training, cross-validation, and generating feature importances.