This Week In Cheminformatics: Issue #025

Handling stochastic polymer representations in databases just got significantly more manageable thanks to a new paper from the Olsen lab on BigSMILES canonicalization. BigSMILES successfully encodes the structural connectivity of complex polymer ensembles, but, its inherent string degeneracy makes routine graph-based searches computationally expensive. The authors present a robust algorithm that maps linear and branched BigSMILES strings onto tree automata state machines that recognize branch points. It then minimizes these branch points into unique graphs, and translates the transition rules back into human-readable, canonicalized strings. Fun read !

Minibaeva et al. present EasyDock 1.3 as a useful tool for large-scale virtual screening campaigns. What makes this release particularly relevant is its complete transition to an open-source ligand preparation pipeline, replacing commercial dependencies with tools like Uni-pKa and MolGpKa for thermodynamic-aware protonation state assignment. Crucially, this pipeline natively integrates PoseBusters and ProLIF for automated post-docking validation and protein-ligand interaction fingerprinting, storing the geometric pass/fail flags and interaction data directly in a SQLite database. I think I’m going to switch to this, goodbye DockingPie, you had a good run.

López-López et al. introduce CRUSH (Cleavage Rules Using SMIRKS Heuristics) as a comprehensive, chemistry-aware bond disconnection strategy. This method applies 33 curated SMIRKS based rules across all eligible bonds at each fragmentation step. The authors benchmarked CRUSH across five diverse datasets including natural products and macrocycles and demonstrated that it consistently shows smaller, Rule of Three-compliant fragments that cover regions of chemical space untouched by existing methods.

• Developing a Machine-Learning Interatomic Potential for Non-Covalent Interactions in Proteins

• SpaceExpander: An Automated System for Drafting Markush Claims to Expand Chemical Space

• Back to the Future of Lead Optimization: Benchmarking Compound Prioritization Strategies

• Accelerating ligand discovery by combining Bayesian optimization with MMGBSA-based binding affinity calculations

• Negative-Ion Mode MALDI-TOF MS Combined with Machine Learning for the Rapid Identification of Colistin-Resistant E. cloacae Complex

• Generative flow model on distance geometry for predicting transition states of chemical reactions

• Targeting the Energy-Coupling Factor Transporters: A Novel Antibiotic Drug Target in Streptococcus pneumoniae

• Deciphering the Molecular Mechanisms of Tau K18 Liquid–Liquid Phase Separation and Its Phosphorylation/RNA-Mediated Modulation via Coarse-Grained Simulations

• Multitask Learning for Membrane Permeability Prediction across Assays with Prediction Reliability Assessment

• CoRE 2D-HOIP DB: Computation-Ready, Experimental Database of Two-Dimensional Hybrid Organic–Inorganic Perovskites

• Development and Validation of CHARMM Force Field Parameters for Major Phytochemicals of Phyllanthus emblica

• Unified Topological Framework for Representation and Construction of Generalized Carbon Nanobelts

• Atomic-level protein–ligand recognition with PBCNet2.0 for probe discovery

• Benchmarking molecular representations and machine learning algorithms for asymmetric catalysis: a palladium-catalysed decarboxylative asymmetric allylic alkylation case study

• OMNI: Optimized Multiview Network Integration with Heterogeneous Graph Attention for Biomedical Interaction Prediction

• Co-Folding–Driven Structural Mapping of Reversible LSD1 Inhibitors Reveals Unexplored Chemical Space and Binding Diversity

• AI-Enforced Ultra-Large Virtual Screening Discovers Potent CD28 Binders

• A Single-Cell Guided Machine Learning Model Predicts Response to Immune Checkpoint Inhibitors in Gastric Cancer

• Band Gap Prediction of Two-Dimensional Materials Using a Gradient-Boosted Feature Selection Approach

• Separation of Flexible Enantiomers Using Shear Flow

• Bridging three-dimensional molecular structures and artificial intelligence with a conformation description language

• MutAIverse: an AI-powered, mechanism-backed platform for discovering novel DNA adducts and their precursor genotoxins

• Predicting enantiomer migration order of levobunolol via sequential computational modeling

• Accelerated Sampling of Protein Dynamics Using BioEmu-Augmented Molecular Simulation

• Discovering CO2–Reactive Carbanions via Property-Guided Generative AI

• The LiPP Benchmark Set for Modeling Lipid–Protein Complexes: Comparison of Co-Folding and Docking Methods

• Structural Characterization of the Type IV Secretion System in Brucella melitensis for Virtual Screening-Based Therapeutic Targeting

• HQMol: Hierarchical Fusion and Query-Guided Alignment for Molecular Graph-Language Modeling

• Reusability report: Assessment of reproducibility and applicability to external datasets for RXNGraphormer

• Determination of bonding radii from small-molecule crystal structures

• Sequence-Derived and Molecular Descriptors for Interpretable Modeling of Molecular Systems: Insights from Peptide Hemolysis

• Synergistic Protein–Protein and Protein–Lipid Interactions Drive SARS-CoV-2 Envelope Assembly

• IPM-DTI: An Interaction-Pattern-Driven Multimodal Framework for Drug–Target Interaction Prediction via Knowledge Graphs

• Librarian of Alexandria: A Modular Chemical Data Extraction Pipeline to Compare LLM Performance

• Navigating molecular OOD-ness

• Comparative evaluation and selection of optimal QSAR-based machine learning model for liver toxicity prediction

• KNexPHENIX: A PHENIX-Based Workflow for Improving Cryo-EM and Crystallographic Structural Models

• Generalizing Drug–Drug Interaction Risk Prediction to Chemically Novel Compounds with Bilinear Interaction Learning and Calibrated Hybrid Inference

• Pushing the Limits of One-Dimensional NMR Spectroscopy for Automated Structure Elucidation Using Artificial Intelligence

• Automating Computational Chemistry Workflows via OpenClaw and Domain-Specific Skills

• Fast Molecular Crystal Structure Prediction Using Sampling by Analogy to Previously Predicted Landscapes

• SpaVGMC: A Unified Representation Learning Framework via Structural and Semantic Alignment in Spatial Transcriptomics

• The Systematic Study of Spatially Conserved Salt Bridges in Protein

• Generative pretraining for drug molecule design with bidirectional structure-property optimization

• Implementation and Validation of Titratable Cysteine in GROMACS-Based Constant-pH Molecular Dynamics

• Structure-Based Discovery of Potent BCL-XL Inhibitors through Rescaffolding

• Structure‐Based Design of Isoxazolidine RIPK1 Inhibitors for Neuroinflammation

• Tracking Gene Expression of Single Mitochondria in Live Neurons Using Nanotweezers

• Burst-Mode Near-Infrared Chemiluminescent Probes for In Vivo Imaging

Ash Jogalekar@curiouswavefn

The ultimate faculty recruitment list: Hermann Weyl recommending physicists and mathematicians for the IAS in 1945. The competition is so stiff that Eugene Wigner and Hans Bethe are considered second-tier.

9:46 PM · Jun 10, 2026 · 33.4K Views

22 Replies · 54 Reposts · 365 Likes

Imran S. Haque (@ihaque@{bsky,genomic}.social)@ImranSHaque

Fable is a psyop to make computational biologists feel like they're real biologists

3:15 PM · Jun 10, 2026 · 551 Views

16 Likes

Keith Robison@OmicsOmicsBlog

Good piece today by @thisischristina on origins of GLP-1 research and how the whole Gila monster saliva exploration was directed at goal of modulating pancreas cells, but I would suggest the following is true

7:23 PM · Jun 9, 2026 · 4.42K Views

2 Replies · 4 Reposts · 64 Likes

Kordel France@kordelfrance

Yes. Problem solving and continual learning do not require trillions of parameters trained on text tokens

Science Magazine @ScienceMagazine

A new Science study shows that bumble bees can position a ball underneath a fake “flower” to reach a reward, suggesting they can exhibit spontaneous problem-solving and challenging the notion that such advanced cognitive abilities are exclusive to large-brained vertebrates.

1:37 AM · Jun 11, 2026 · 64K Views

14 Replies · 33 Reposts · 1.54K Likes

Simo Ryu@cloneofsimo

5:37 AM · Jun 10, 2026 · 42.2K Views

10 Replies · 90 Reposts · 1.4K Likes

Ash Jogalekar@curiouswavefn

From the immortal Richard Hamming.

4:29 PM · Jun 9, 2026 · 3.07K Views

1 Reply · 10 Reposts · 76 Likes

Patrick Boyle — e/🦀@p_maverick_b

AI labs wanting to solve biology and learning how our data is organized

12:36 PM · Jun 9, 2026 · 58K Views

18 Replies · 75 Reposts · 887 Likes

Sebastian S. Cocioba🪄🌷@ATinyGreenCell

Because biology is real

Anthropic @AnthropicAI

New Science Blog: Why has AI advanced faster in coding than in biology? To agents, bio databases are like cities built before cars—maddening to drive in because they're designed for different traffic. How do we build infrastructure agents can use? https://t.co/PQaNQ4GRJZ

10:20 PM · Jun 8, 2026 · 31.6K Views

9 Replies · 21 Reposts · 394 Likes

Paul Gavrikov@PaulGavrikov

Made me laugh way to hard #CVPR2026

4:07 PM · Jun 8, 2026 · 33K Views

2 Replies · 16 Reposts · 300 Likes

stay hydrated,
Manas