rewire.it
Published evidence

Results from the literature

Numerical benchmark results as their papers report them, with the dataset, metric and source table beside every number.

Paper-reported results. These are not rewire.it runs. Scores from different datasets, splits or protocols are not a single leaderboard.

100 verified papers 149 numerical rows

Results CSV →Papers JSON →

149 of 149 paper-reported rows match · 50 displayed

Paper-reportedCells & tissuesPeer-reviewed2026

Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing

Best frozen single-cell foundation model Donor-aware age-class prediction

Reported score
0.322 ± 0.008 standard deviation
Metric
Balanced accuracy
Dataset / split
AIDA v2 PBMC cohort 622 donors

Same donor-aware splits and logistic-regression probe as expression PCA; text names Geneformer as best model on AIDA v2.

Independent paper evaluationTable 2, AIDA v2 row, scFM BA ± SD column →
Paper-reportedProteins & complexesPeer-reviewed2026

PRIME: An evaluation framework for protein representation inference and generalization in viral mutation space

ESM-2 (8M) Mutated RBD binding prediction

Reported score
0.0248 ± 0.01
Metric
R²
Dataset / split
PRIME mutated RBD · position-stratified

Frozen mean-pooled representation with downstream regression; position-stratified split.

Independent paper evaluationTable 1, ESM-2 8M / Mean / not fine-tuned row, Position-Stratified Split Binding R² column →
Paper-reportedProteins & complexesPeer-reviewed2026

PRIME: An evaluation framework for protein representation inference and generalization in viral mutation space

ESM-C (300M) Mutated RBD binding prediction

Reported score
-0.0162 ± 0.01
Metric
R²
Dataset / split
PRIME mutated RBD · position-stratified

Frozen mean-pooled representation with downstream regression; position-stratified split.

Independent paper evaluationTable 1, ESM-C 300M / Mean / not fine-tuned row, Position-Stratified Split Binding R² column →
Paper-reportedMolecular interactionsPeer-reviewed2025

A Comparative Study of Deep Learning and Classical Modeling Approaches for Protein–Ligand Binding Pose and Affinity Prediction in Coronavirus Main Proteases

Boltz-2 Ligand potency prediction using generated poses

Reported score
0.800 ± 0.027
Metric
Pearson R
Dataset / split
SARS-CoV-2 Mpro ligands

Potency prediction using Boltz-2 ligand-pose generation protocol; see paper scoring pipeline.

Independent paper evaluationTable 3, Boltz-2 row, Pearson’s R column →
Paper-reportedMolecular interactionsPeer-reviewed2025

A Comparative Study of Deep Learning and Classical Modeling Approaches for Protein–Ligand Binding Pose and Affinity Prediction in Coronavirus Main Proteases

DiffDock Ligand potency prediction using generated poses

Reported score
0.695 ± 0.037
Metric
Pearson R
Dataset / split
SARS-CoV-2 Mpro ligands

Potency prediction using DiffDock ligand-pose generation plus paper scoring pipeline; not a native DiffDock affinity score.

Independent paper evaluationTable 3, DiffDock row, Pearson’s R column →
Paper-reportedCells & tissuesPeer-reviewed2025

Cell-DINO: Self-supervised image-based embeddings for cell fluorescent microscopy

Cell-DINO ViT-L protein localization classification

Reported score
65.5%
Metric
F1
Dataset / split
HPA-FoV

Self-supervised microscopy embedding pre-trained on HPA-FoV; downstream protein-localization classifier. Dataset-specific pretraining; the paper does not claim a general-purpose foundation model that generalizes beyond these benchmarks.

Author's modelTable 2, HPA-FoV section, Cell-DINO row, PL column →