Metagenomics

EMG produced TPA metagenomics assembly of the PRJNA300541 data set (Antibiotic resistance exchange between microbiota in resource-poor settings in Latin America).

Microbiome sequencing from a study of antibiotic resistance exchange in resource-poor Latin American settings, reassembled by MGnify. The benchmark uses InterPro functional-abundance profiles to predict the recorded environmental biome of each sample.

Binary classificationLicence: PublicMGYS00005381Target: environment (biome)mgnify ↗

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
TabPFN 3.5	0.9507	0.9109	0.9536	0.9827
Kumo Tabular (M)	0.946	0.8959	0.952	0.9803
LimiX2	0.9415	0.8867	0.9487	0.9867
RealTabPFN 2.5	0.941	0.8932	0.9419	0.9824
Logistic Regression	0.9384	0.8881	0.9421	0.9734
Logistic Regression (tuned)	0.9384	0.8881	0.9421	0.9729
Causilo	0.9324	0.8767	0.9395	0.9799
TabPFN Wide 5k (ne3)	0.9323	0.8805	0.9335	0.9778
AutoGluon	0.9315	0.87	0.936	0.9712
TabDPT	0.9315	0.87	0.936	0.972
MLP (tuned)	0.9304	0.8776	0.9382	0.9716
MLP	0.9288	0.8704	0.9303	0.9572
TabFM	0.9288	0.8704	0.9303	0.9664
RealTabPFN v2	0.9245	0.865	0.9221	0.9656
RealMLP	0.9194	0.8502	0.9235	0.9619
Extra Trees	0.9125	0.8504	0.9134	0.9427
Extra Trees (tuned)	0.907	0.8417	0.9076	0.9373
Random Forest	0.9058	0.8294	0.907	0.9402
MITRA	0.901	0.8221	0.9075	0.9632
XGBoost (tuned)	0.8997	0.8284	0.9045	0.9467
Random Forest (tuned)	0.895	0.8242	0.8958	0.9342
GeneICL	0.8888	0.7846	0.8985	0.9357
TabM	0.8825	0.7883	0.8924	0.9079
XGBoost	0.8633	0.7638	0.8562	0.9241