AI for Bio · Model

KERMT

Graph-transformer chemical pretrained model with multitask fine-tuning and acceleration for in-house ADMET prediction

Small moleculesMixedUnited StatesActive中文

Overview

NameKERMT · aliases: KERMT, Contrastive KERMT, NV-KERMT-70M-v2
Full nameKinetic GROVER Multi-Task
Ambiguous names (not auto-linked)None
Familyv1 and the contrastive v2.0 are two papers with separate weights; results must not be carried across
  • KERMT v1 · †
  • Contrastive KERMT v2.0 ·
OrganizationMerck & Co. (modeling and informatics), NVIDIA (BioNeMo team)
Organization typeMixed
CountryUnited States
Model classProperty prediction (graph-transformer pretraining + multitask fine-tuning)
DomainSmall molecules
First release† partly verified · paper revision
Basis: First preprint (arXiv 2510.12719)
Note: The arXiv id and date are carried over from the site record and were not re-fetched this round source
Other milestones
  • First public evidence: † · First preprint source
  • Preprint v1: † · First preprint source
  • Code release: · GitHub v1.0.1 (v1.0.0 earlier) source
  • Weights release: · v2 weights uploaded to the model hosting platform source
Latest update · new version
Basis: GitHub release v2.0.0 (contrastive version)
Contrastive v2.0: code release v2.0.0, v2 weights uploaded 2026-06-09 with a matching preprint (2026-06-09); commits through 2026-09-15 are documentation and tooling maintenance source
The most recent sourced event, not a claim that nothing changed since.
StatusActive · New version, new weights and a new preprint within 12 months
OpennessCode: Open · Apache-2.0
Weights: Research use only · NVIDIA Open Model License (v2.0 weights)
Code is open; v2.0 weights are offered on a model hosting platform under an open model licence, while the pharma company's own page states no licence
Benchmarks reported0 counted · none of the 44
Wet-lab validationNone recorded
Unverified fields† first_release, release_milestones.paper_v1, release_milestones.first_public_evidence
As of2026-11-16

What it does

Adapts a third-party graph-transformer chemical pretraining architecture - atom and bond message passing plus a transformer - pretrained on tens of millions of molecules from two public compound libraries, with the emphasis on multitask fine-tuning and hardware-side acceleration and memory optimisation. It is positioned as a production ADMET model inside a pharma company, compared against the multitask variant of Chemprop and sitting on the same pretrained-ADMET track as MiniMol and MolE.

Benchmarks reported

None of the 44 benchmarks. This model is on the zero-pairing list.

How it evaluates itself: Merck internal ADMET (30 endpoints); Public literature ADMET (25 endpoints) and Biogen set

Other evaluations (outside the 44)

  • In-house pharma ADMET endpoint set (time split) (Own task set) Self-reported: A time-split evaluation over 30 endpoints and about 800k compounds against a multitask graph network and another pretrained model; the data are not public and cannot be reproduced externally
  • Own public-literature ADMET set and a third-party public set (Own task set) Self-reported: 25 public endpoints plus a six-endpoint public set; in the same paper these two reverse the ordering seen on the in-house set, so the ranking depends on the setting
  • Similarity-binned generalisation analysis (Comparison against baselines only) Self-reported: Comparison by chemical-similarity bin, with the authors reporting an advantage over the multitask graph-network baseline in every bin; all self-reported

Developers report: multitask fine-tuning clearly beats non-pretrained graph-network baselines, with the advantage growing at larger data sizes, plus an accelerated implementation.

Independent evidence: No independent evidence found. No site-benchmark pairing and no third-party evaluation found; the key results rest on non-public in-house data

Training data

Group in Figure 7: Private or licensed data.

ClassValueDetail
a. Public databasesYes25 public ADMET endpoints with 114,112 compounds plus another public set of 6 endpoints and 3,521 compounds; pretraining on 11 million compounds from two public libraries source
b. Private or licensed dataYesIn-house pharma ADMET data: 30 endpoints, 800,733 compounds measured through 2024, not public source
c. Own wet-lab data used for trainingNoThe in-house data are the company's historical measurements, not generated for this model source
d. Wet lab used only for validationNo

Wet-lab validation

None. The developers report no wet-lab validation.

Background

  • Organization: A modelling and informatics department at a large US pharma company co-authoring with the biological-models team of a GPU company; compute and funding come from both companies' internal R&D budgets, and the model is also listed in the GPU company's open-model catalogue.
  • Funding: Not disclosed
  • Products and partnerships: The architecture comes from an open graph-transformer pretraining model published by a Chinese internet company's lab; the code is open under the GPU company's digital-biology GitHub organisation and the pharma company published a corporate story page (2026-03-19).
  • Related models: Chemprop v2, MiniMol, MolE, GenMol

Reading notes

  1. The most informative results come from the in-house dataset, which is not public and cannot be reproduced or checked externally.
  2. There are no site-benchmark results, so it cannot be compared directly with ADMET-AI or MiniMol.
  3. The advantage over graph-network baselines holds only in the multitask setting on the in-house assays; the same paper's public and third-party sets reverse the order, and all of it is self-reported.
  4. The TDC pairing does not hold: that benchmark does not appear in the paper.

Caveats

  1. Zero pairings: no results on any site benchmark.
  2. The key comparison depends on non-public in-house pharma data.
  3. v1 and the contrastive v2.0 are different weights; results cannot be carried across.
  4. The first-release date is carried over from the site record and was not re-fetched this round.

Sources

  1. First preprint (preprint)
  2. First preprint full text (preprint)
  3. Contrastive-version preprint (preprint)
  4. Code release list (repository)
  5. v2 weights page (package registry)

Corrections

2 corrections
FieldWasNowBasisFound
benchmarks_reportedCandidate pairing TDCNot a pairing: the benchmark does not appear in the paper; the actual benchmarks are an internal set and two public setssource2026-11-16
latest_update2025-10 preprint2026-06-10 contrastive-learning version v2.0 (code, weights and new preprint)source2026-11-16

Part of the AI for Bio benchmarks survey · all 112 models · open in the explorer. Record as of 2026-11-16.