Cancer Diagnosis and Treatment

This study uses ctDNA profiling to predict tissue-of-origin and identify actionable biomarkers in cancer of unknown primary, providing diagnostic and therapeutic insights in most cases.

TCUP predicts tissue of origin and cancer of unknown primary using ensemble learning with embeddings and ablation analysis, achieving high accuracy and providing gene importance insights.

This paper develops OncoOrigin, a machine-learning tool that accurately predicts primary cancer sites from tumor DNA sequencing data in CUP patients, achieving high top-2 accuracy and ROC-AUC.

This paper develops OncoNPC, a machine learning classifier for CUP patients using tumor sequencing data, showing improved survival outcomes with treatment aligned to predicted cancer types.

The paper develops a random forest classifier using genome-wide mutation features to classify cancer of unknown primary (CUP) with high accuracy.

This study combines DNA sequencing and gene-expression profiling with clinical data to improve diagnosis of cancer of unknown primary (CUP), identifying potential tissue origins in over a third of cases.

TRANSCUP is a software tool for predicting cancer of unknown primary using transcriptome data, achieving high accuracy in tumor classification.

This paper develops machine learning models to predict the primary site of origin and molecular subtype of metastatic cancers using gene expression data from TCGA, achieving high accuracy across multiple cancer types.