AI Science Tracker

Numerion Labs (United States)

2 discoveries from 2020 to 2024

Discoveries

NGT: Generative AI with Synthesizability Guarantees Discovers MC2R Inhibitors from a Tera-Scale Virtual Screen.

Source check: Needs source check. Not independent scientific confirmation.

Finding / tested result: The primary abstract reports MC2R inhibitors; ledger-recorded publisher supporting-information descriptions identify cell-based antagonist concentration-response assays. Supporting data were not directly read.

Technology: NeuralGenThesis generative model and normalizing flow with reinforcement learning

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
NeuralGenThesis generative model and normalizing flow with reinforcement learning is the AI method documented in the editorial source review.
Limitation / still unknown
The date refers to the cited version. Earlier public versions and independent replication were not systematically verified.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://doi.org/10.1021/acs.jmedchem.4c01763

Needs source check · editorial details

The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review scope and date are not recorded; source checked does not establish full-text review.

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Discovery of Novel Inhibitors of a Critical Brain Enzyme Using a Homology Model and a Deep Convolutional Neural Network.

Source check: Needs source check. Not independent scientific confirmation.

Finding: AtomNet, a deep convolutional network, helped identify five low-micromolar ANAT enzyme inhibitors using a homology model despite the lack of a native structure or high-identity template. This is biochemical hit identification for a Canavan-disease target, not evidence of treatment efficacy.

Technology: AtomNet deep convolutional neural network

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
AtomNet, a deep convolutional neural network, ranked candidate ligands for ANAT using a homology model and informed biochemical hit selection.
Test and result
The primary abstract reports five low-micromolar ANAT inhibitors. Indexed ACS text describes dose-response validation and comparison of AtomNet and smina rankings over a 7.2-million-compound library; full assay denominator and comparator details were not directly reviewed.
Limitation / still unknown
Full ACS Methods and supporting information were not reviewed. Exact assay denominator, comparator details, cellular target engagement, in-vivo efficacy and therapeutic safety were not assessed; biochemical inhibition does not establish treatment efficacy.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://doi.org/10.1021/acs.jmedchem.0c00473

Needs source check · editorial details

The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 18 Nov 2026. Review scope: Indexed primary PubMed abstract and publication metadata; full ACS Methods/supporting information not reviewed..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open the interactive profile for the timeline, or browse all organizations and model families.