Metagenomics

EMG produced TPA metagenomics assembly of PRJNA728374 data set (Gut microbiome responses to dietary intervention with hypocholesterolemic vegetable oils.).

Gut microbiome sequencing from a dietary intervention study involving cholesterol-lowering vegetable oils, reassembled by MGnify. The benchmark uses InterPro functional-abundance profiles of microbial communities to predict recorded host sex.

Binary classificationLicence: PublicMGYS00006282Target: host sexmgnify ↗

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
GeneICL	0.6402	0.3005	0.6455	0.6253
Logistic Regression (tuned)	0.6127	0.2409	0.6177	0.5764
TabFM	0.6057	0.24	0.6138	0.6238
Logistic Regression	0.6003	0.2174	0.6063	0.5889
TabPFN 3.5	0.5982	0.2178	0.6052	0.6073
RealTabPFN 2.5	0.5972	0.2163	0.6058	0.5914
AutoGluon	0.5891	0.2239	0.604	0.5782
TabDPT	0.5771	0.1693	0.5829	0.5851
Kumo Tabular (M)	0.5724	0.1698	0.5818	0.5892
MLP (tuned)	0.5654	0.1402	0.5688	0.5713
Causilo	0.5572	0.139	0.5648	0.5775
TabPFN Wide 5k (ne3)	0.5543	0.15	0.5663	0.6001
RealMLP	0.5393	0.1651	0.5715	0.6224
MLP	0.5388	0.08975	0.5437	0.5527
Extra Trees	0.5337	0.08199	0.5389	0.5358
MITRA	0.5334	0.08474	0.5411	0.5286
RealTabPFN v2	0.5309	0.07876	0.5388	0.5429
TabM	0.5245	0.06697	0.5322	0.5315
XGBoost	0.5143	0.03866	0.5183	0.5151
Extra Trees (tuned)	0.5086	0.03366	0.5155	0.5022
XGBoost (tuned)	0.5072	0.03264	0.5162	0.5135
Random Forest	0.504	0.03702	0.5163	0.521
Random Forest (tuned)	0.4892	0.004444	0.5006	0.5229
LimiX2	0.4149	-0.163	0.4212	0.4004