Metagenomics

EMG produced TPA metagenomics assembly of the A metagenome-wide association study of gut microbiota in type 2 diabetes. (gut metagenome) data set.

Gut microbiome data from a study of type 2 diabetes, reassembled and annotated by MGnify. The benchmark uses InterPro functional-abundance profiles of the microbial community to predict recorded host sex.

Binary classificationLicence: PublicMGYS00005285Target: 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
Kumo Tabular (M)	0.5411	0.09658	0.5442	0.5722
AutoGluon	0.5156	0.06304	0.5285	0.5112
TabDPT	0.5078	0.0498	0.5199	0.5181
Logistic Regression (tuned)	0.502	0.02194	0.5086	0.4852
Logistic Regression	0.4937	0.004989	0.5012	0.4869
TabFM	0.4875	0.038	0.5143	0.4916
MLP	0.4842	-0.01412	0.493	0.5068
MLP (tuned)	0.4831	-0.02218	0.4891	0.4807
RealTabPFN 2.5	0.4812	-0.01983	0.4886	0.5063
GeneICL	0.4811	0.004206	0.5004	0.5057
XGBoost	0.4798	-0.02277	0.4919	0.5083
Causilo	0.4636	-0.04294	0.4832	0.4593
MITRA	0.4632	0.02605	0.5097	0.4861
Extra Trees (tuned)	0.4539	-0.05042	0.4772	0.4518
RealMLP	0.4457	-0.0659	0.469	0.4811
Random Forest (tuned)	0.4455	-0.05942	0.4716	0.4613
RealTabPFN v2	0.4438	-0.02179	0.491	0.4843
TabPFN 3.5	0.4411	-0.02343	0.4859	0.4831
Random Forest	0.4355	-0.08029	0.4681	0.4597
Extra Trees	0.4339	-0.07003	0.4704	0.4614
TabM	0.4323	-0.07664	0.4563	0.4533
LimiX2	0.4295	0.008401	0.5105	0.4731
XGBoost (tuned)	0.4231	-0.05529	0.4756	0.4767
TabPFN Wide 5k (ne3)	0.3716	0	0.5	0.4572