Probabilistic Contrastive Pretraining Framework Improves ADME Property Prediction for Drug Discovery
Researchers have proposed a molecular graph-transformer pretraining framework called Contrastive KERMT that improves prediction of ADME drug properties by combining contrastive learning with chemistry-specific self-supervision. ADME properties—absorption, distribution, metabolism, and excretion—are critical benchmarks in drug discovery but are difficult to model due to noisy, interdependent, and scarce data. The method demonstrates consistent performance gains of roughly 7–10% over a prior baseline across three datasets, potentially accelerating early-stage drug development.
A preprint posted to arXiv introduces Contrastive KERMT, a pretraining approach for predicting ADME properties of drug candidates using a molecular graph-transformer architecture. The framework encodes molecular graphs into latent variables, reconstructs SMILES chemical notation strings from those latent codes, and integrates contrastive mutual information learning (cMIM) alongside domain-specific chemistry tasks. Crucially, rather than treating these objectives as separately weighted auxiliary losses, the authors unify reconstruction, contrastive discrimination, and chemistry supervision as unit-weighted log-probability factors within a single probabilistic latent-variable objective. For downstream fine-tuning, a multi-task graph neural network readout architecture with task-specific multilayer perceptron heads is used to reduce negative transfer across heterogeneous ADME endpoints. Evaluated on three datasets—Biogen, ExpansionRX, and ChEMBL-MT—the method improves over the KERMT baseline by 7.6%, 9.9%, and 9.5% respectively on significantly improved endpoints. The authors also find that expanding the pretraining corpus with ADME-adjacent molecules further boosts transfer performance, and that the contrastive component produces more chemically coherent latent representations.
What's missing
Whether the reported improvements hold across all ADME endpoints or only a subset of 'significantly improved' ones, and how the method compares to other recent graph-based or large language model baselines beyond KERMT. The computational cost and data requirements for pretraining are not discussed in the abstract.
What different sources said
- arXiv cs.LGCenter
Probabilistic Contrastive Pretraining for Multi-task ADME Property Prediction
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.