Investors

Investors

A

task-conditioned

deep

deep

learning

learning

deep

learning

model

for

Ames

mutagenicity

prediction

The highest out-of-domain sensitivity of any model evaluated, with no trade-off in accuracy.

Read the Paper

Best-in-Class

AmesNet™ achieves a sensitivity of 0.72 and a balanced accuracy of 0.81 on out-of-domain test data. A sensitivity improvement of up to 33% over the leading models.

Best-in-Class

AmesNet™ achieves a sensitivity of 0.72 and a balanced accuracy of 0.81 on out-of-domain test data. A sensitivity improvement of up to 33% over the leading models.

New Model Paradigm

Task-Conditioned Learning (TCL) architecture captures assay-context-dependent mutagenicity that all prior “Unconditioned” models either don’t account.

New Model Paradigm

Task-Conditioned Learning (TCL) architecture captures assay-context-dependent mutagenicity that all prior “Unconditioned” models either don’t account.

Early-Stage Screening

AmesNet’s performance enables high-confidence Ames screening at scale during early discovery, pulling preclinical safety assessment forward.

Early-Stage Screening

AmesNet’s performance enables high-confidence Ames screening at scale during early discovery, pulling preclinical safety assessment forward.

Contact

Put AmesNet to Work on Your Chemistry

Evaluate AmesNet on compounds from your own discovery or development program.

AmesNet is Model Medicines’ task-conditioned deep learning model for Ames mutagenicity prediction, designed to identify potential liabilities earlier and perform reliably in novel chemical space.

We are opening a limited number of AmesNet Pilot engagements for biopharma and drug discovery teams interested in evaluating the model against their own chemistry.

Apply for an AmesNet Pilot