This Week In Cheminformatics: Issue #034

Partially. The best ”agreeing” model, Gemini 3.1 Pro, reached MCC of ~0.4 against human expert consensus, while Claude Opus 4.8 (0.2), GPT-5.5 (0.35), and Llama 3.1 70B (0.07) showed weaker and less reliable agreement. Betteridge’s Law of Headlines almost holds.

SHAP (SHapley Additive exPlanations) assigns each input feature a number showing how much it changes a specific prediction, using Shapley values from game theory. Shapley values do this by looking at every possible subset of features, checking what the model predicts with and without a given feature added to that subset, and averaging that difference across all subsets. This average (marginal) contribution is the feature’s Shapley value. This means with N features, there are 2^N possible subsets i.e. for a 2048-bit fingerprint, there are 2²⁰⁴⁸ possible subsets. That’s way too many calculations. So, there are approximations to it such as kernalSHAP, treeSHAP, etc. It’s obviously not perfect with many approximations such as assuming feature independence, etc. read more here. In this paper, Roth redefined the “players” in the Shapley value game to be resulting BRICS fragments instead of fingerprint bits or graph nodes, which makes exact Shapley value computation feasible by exhaustively evaluating the model on all “fragment subsets”. They applied this to RF and GCN models trained on aqueous solubility, Ames mutagenicity, and SARS-CoV-2 Mpro potency and compared the resulting fragment-level attributions against standard SHAP and GNNExplainer. They show Fragment-level Shapley values recovered known chemical trends across all three tasks. Compared to SHAP on bits and GNNExplainer, the fragment-based attributions were seemingly more chemically meaningful, though the two models (RF vs. GCN) differed somewhat in exact fragment rankings.

Sayyah et al. screened FDA-approved drugs against BCL-2 and found relugolix as a hit. I did a report on BCL-2 and venetoclax as the BH3 mimetic for my medchem class last year, so I remember that venetoclax is no ordinary drug from a design, synthesis and structural perspective and if you don’t know the FBDD story behind it, you need to check out the amazing https://fbdd-lit.blogspot.com and https://practicalfragments.blogspot.com and papers like this one. So, when I see a repurposing claim it’s very interesting, and even better for the nerds, this paper does in vitro studies to back up the claim. The authors used NeuralPlexer to predict BCL-2–ligand complex structures for 3094 FDA-approved drugs, then filtered candidates through docking refinement, MD simulations, and more to compare each candidate’s dynamic interaction pattern to the reference inhibitor, Venetoclax. This identified relugolix (oral drug primarily used to treat advanced prostate cancer or heavy menstrual bleeding due to uterine fibroids) as a promising BCL-2 hit, which they then confirmed experimentally via TR-FRET binding and cancer cell viability assays. Pretty cool !

• MMAllo: a multimodal deep learning framework and GaMD simulations for predicting protein allosteric sites and allosteric mutations

• Molecular dynamics study of composition-dependent intermolecular interactions and trigger-bond stability in DNTF/HMX melt-cast explosive mixtures

• Q² performance: Toward Unbiased Selection of Machine-Learning Regression Models

• Theoretical insights into the free radical scavenging mechanisms and metal chelating ability of lobelialsine: a DFT study

• Thorough study of the properties of the half-Heusler compounds TbPdBi and ErPdBi employing the ab initio method with GGA and GGA+U formalisms

• MolScope: A Transparent Representation-Aware Workflow Toolkit for Reproducible Medicinal-Chemistry Triage

• Comparative Assessment of Similarity Metrics for 1H NMR Spectral Matching: Evidence for the Robustness of the Normalized Square Dot Product

• Synthetic data for more accurate deep learning models in molecular science: a test case of protein-ligand binding affinity prediction

• Reinforcement learning for de novo RET inhibitor design: potency-focused optimization and polypharmacological multi-parameter optimization against resistance mutations

• Robust generative transition-state models for unseen chemistry

• Computational techniques to study breast cancer scaffolds for antiangiogenesis: a review

• Organometallic-enhanced chromone-thiazole hybrid antioxidants: a DFT study of solvent-dependent HAT/SPLET switching

• Development and validation of a ReaxFF reactive force field for the pyrolysis of fluorine-containing gas systems

• Theoretical analysis on the structural, electronic, and intermolecular interaction features of the CL20/1-AMTN cocrystal

• Nanomechanical response of peptide-bound DNA as a model molecular bio-nanocomposite: a steered molecular dynamics study

• Spatial-spectral fusion enables drug repositioning by capturing indirect and long-range associations in biological networks

• B-MASTER: scalable Bayesian multivariate regression for master predictor discovery in colorectal cancer microbiome-metabolite profiles

• Sequencing saturation does not uniquely determine molecular recovery in UMI transcriptomics

• Phylogenetic inference under the balanced minimum evolution criterion via semidefinite programming

• ConfRetro: a 3D-aware template-free method for enhancing retrosynthesis via molecular conformer information

• Assessing the influence of different alignment tools on the accuracy of a forensic epigenetic clock

• Unraveling the R-Enantiomer Preference in Mirror-Image Cyclodextrin: An Integrated Multiscale Computational Study

• MMGNN: Multilevel, Multicolor Graph Neural Networks for Molecular Property Prediction

• Unraveling the R-Enantiomer Preference in Mirror-Image Cyclodextrin: An Integrated Multiscale Computational Study

• Extending the Python RESP Framework for Extra Point Charge Fitting

• Methodological Considerations in Small-Sample Multitask QSPR Modeling for PAMPA Permeability

• MMGNN: Multilevel, Multicolor Graph Neural Networks for Molecular Property Prediction

• In Silico Decoding of Nonsteroidal Mineralocorticoid Receptor Antagonist Binding: Implications for Drug Design

• Differentiable Thermodynamic Phase-Equilibria for Machine Learning

• Ser715 Phosphorylation Induces β-Hairpin Unfolding and Destabilization of FOXM1 Autoinhibition

• Dual-Attention Multimodal Framework for Molecular Property Prediction

• Machine-Learning-Guided Design of Antifreezing Peptides

• A Kinetic Origin of Vps34-IN-3 Selectivity for VPS34 over PI3Kα Revealed by Supervised Molecular Dynamics and Potential of Mean Force Analysis

• FastRet: Fast and Simple Retention Time Prediction in Liquid Chromatography

• Systematic Construction of Chemical Reaction Networks: An Automated Nanoreactor Approach Powered by Integrated Tempering Sampling

• Accurate Identification of Covalently Ligandable Cysteines Using CCSite

• AI-Guided Discovery of Natural-Product-Inspired Scaffolds as Dual P-gp/BCRP Inhibitors to Overcome Cancer Multidrug Resistance

• Chemical Stabilization and Transformation of Schiff Base Scaffold: A

  Path toward Superior Platinum-Based Anti-Cancer Stem Cell Drugs

• Structure and Dynamics of the HIV‑1 Envelope Protein on the Virion

  Envelope

• How Polymer Chains Remember Their Crystalline Past: Simulations

  Uncover the Molecular Origin of Melt Memory

• Protein Modifications for Cellular Protein Delivery

• New Frontiers in Single-Atom Catalysis: Active-Site Dynamics and AI-Enabled Design

• Practical Crystallization Strategies for Medicinal Chemists

bomberfish 🍁@bomberfish77

this is why they keep asking you to donate btw

System Settings @app_settings

“oh man, i love using the internet archive! it’s so usefu–”

11:38 PM · Aug 15, 2026 · 298K Views

15 Replies · 1.54K Reposts · 26K Likes

Haniel Ulises@Haniel_Ulises

This book is a gem released early this year. By Victor Eijkhout, the mind behind "The Art of HPC" book series. If you love modern c++ and HPC this is a must

5:37 AM · Aug 16, 2026 · 21.8K Views

6 Replies · 57 Reposts · 791 Likes

DROID@droidbuilds

9:11 AM · Aug 16, 2026 · 212K Views

65 Replies · 838 Reposts · 14.2K Likes

ren ᶻ𝗓𐰁@acorpseofagirl

i love how everyone still calls twitter twitter

2:24 PM · Aug 14, 2026 · 13.6M Views

2.24K Replies · 85.8K Reposts · 864K Likes

Holly Blake@hollycurates

hope you guys made it to the annual reports side of twitter this week, because this post sent me down the best rabbit hole possible!!

Oliver ೫ @oliverhamrin

Four of my favorite annual report covers. If I had to single out one year, I’d say this lost art peaked in 1977. https://t.co/vJU2Mnc2MU

1:56 AM · Aug 16, 2026 · 104K Views

20 Replies · 322 Reposts · 4.39K Likes

Jean P.D. Meijer ― 🇪🇺 eu/acc@initjean

"87,605 branches"??? this was the guy taking down GitHub all the time

Tomas Reimers @TomasReimers

Good weekend to get sOme rest

10:22 AM · Aug 15, 2026 · 172K Views

34 Replies · 90 Reposts · 4.72K Likes

Boring_Business@BoringBiz_

4:06 PM · Aug 15, 2026 · 315K Views

17 Replies · 1.14K Reposts · 16.6K Likes

Ash Jogalekar@curiouswavefn

Leo Szilard was known for taking long walks during which he just thought. So was Paul Dirac. So were many British thinkers in particular, for instance David Hume and Adam Smith. My dad who was both a walker and a thinker (he was a college professor) always cited these examples to

Ihtesham Ali @ihteshamali

A Stanford psychologist spent 4 years proving that the simple act of walking generates 60% more creative ideas than sitting, and the experiment she designed to kill every alternative explanation is one of the most decisive findings in modern psychology. Her name is Marily

10:38 PM · May 25, 2026 · 29.7K Views

25 Replies · 42 Reposts · 381 Likes

Blank@blankparticle

🫩

9:40 PM · Aug 15, 2026 · 13.5K Views

7 Replies · 1 Repost · 273 Likes

Oliver ೫@oliverhamrin

Four of my favorite annual report covers. If I had to single out one year, I’d say this lost art peaked in 1977.

TJ Pratt @TJ_Pratt

I’ve fallen down the vintage annual report design rabbit hole https://t.co/yH4uHQQKKN

9:25 PM · Aug 14, 2026 · 427K Views

35 Replies · 690 Reposts · 9.32K Likes

no context memes@nocontextmemes

10:19 PM · Aug 15, 2026 · 172K Views

42 Replies · 1.69K Reposts · 16.8K Likes

owl@owl_posting

more biology research organizations should consider running competitions for problems they view as neglected. its an incredible social good that also happens to be beneficial for everyone involved: 1. the sponsoring institution gets to act as a tastemaker, helping shape the

7:25 PM · Aug 14, 2026 · 11.4K Views

18 Replies · 26 Reposts · 233 Likes

Best,
Manas