Gene expression

TCGA Breast (BRCA, RNA-seq)

RNA sequencing measures gene expression in breast tissue samples from The Cancer Genome Atlas breast cancer project. The benchmark uses normalized expression values to predict whether each sample is labelled as tumor or normal tissue.

Binary classificationLicence: NIH Genomic Data Sharing Policy (GDC open-access)TCGA-TCGA-BRCA_Gene-Expression-QuantificationTarget: sample_typetcga ↗

At this operating point

Model comparison

p=10,000, n=100 · Higher Macro-F1 is better

Random Forest is marked with a diamond. Models are ordered by the selected metric.

Across sample budgets

Learning curves

Models

All metric values at this operating point

Model	Macro-F1	MCC	Balanced accuracy	ROC-AUC
RealTabPFN v2	0.976	0.9528	0.9836	0.999
RealTabPFN 2.5	0.9733	0.9474	0.9756	0.9989
TabPFN Wide 5k (ne3)	0.9706	0.9424	0.971	0.999
Kumo Tabular (M)	0.9691	0.9394	0.9784	0.9988
TabPFN 3.5	0.9684	0.9376	0.9706	0.999
LimiX2	0.9574	0.9182	0.9876	0.9984
RealMLP	0.9566	0.9157	0.9493	0.9977
GeneICL	0.9449	0.8956	0.9847	0.9979
MLP	0.9439	0.8893	0.9388	0.9953
AutoGluon	0.9427	0.8866	0.9341	0.9951
TabDPT	0.9418	0.8856	0.9458	0.9942
Causilo	0.9393	0.8838	0.915	0.9974
MITRA	0.931	0.8665	0.9126	0.9957
TabM	0.9177	0.8429	0.9319	0.995
Random Forest	0.9145	0.8406	0.8759	0.9976
Logistic Regression	0.8951	0.802	0.8568	0.994
Random Forest (tuned)	0.8535	0.7197	0.8411	0.8987
Extra Trees	0.85	0.734	0.792	0.9962
MLP (tuned)	0.8412	0.6925	0.8469	0.8953
XGBoost	0.8402	0.7109	0.8223	0.9826
Extra Trees (tuned)	0.8209	0.664	0.7979	0.8985
Logistic Regression (tuned)	0.7771	0.5829	0.7426	0.895
XGBoost (tuned)	0.7726	0.583	0.79	0.8954
TabFM	0.6765	0.4384	0.632	0.9962