Optimization of binding affinities in chemical space for drug discovery
Bibliographic record
Abstract
Drug discovery strategies can be broadly categorized into ligand-based approaches, that leverage molecules with observed bioactivity, and structure-based ones, where drug candidates are specifically designed to interact with a target receptor in the human body.Ligand-based drug design has recently benefited from the development of deep generative models.These models enable extensive explorations of the chemical space, and provide a platform for molecular optimization.However, the vast majority of current methods do not leverage the structure of the binding target, which potentiates the binding of small molecules and play a key role in the interaction.We propose an optimization pipeline that leverages complementary structure-based and ligand-based methods.Instead of performing docking on a fixed chemical library, we iteratively select promising compounds in the full chemical space using a ligand-centered generative model.Molecular docking is then used as an oracle to guide compound optimization.This allows to iteratively generate compounds that fit the target structure better and better, without prior knowledge about bio-actives.For this purpose, we introduce a new graph to selfies Variational Autoencoder (VAE) which benefits from an eighteen times faster decoding than the graph to graph state-of-theart, while achieving similar performance.We then successfully optimize the generation of molecules towards high docking scores, enabling a ten-fold enrichment of high-scoring compounds found with a fixed computational cost.i AbrgTraditionnellement, les problmes de drug design, qui consistent proposer de nouveaux ligands pour des rcepteurs protiques connus, sont abords selon deux approches: Soit en partant de la structure 3D du rcepteur et en recherchant les ligands les plus complmentaires dans une librairie, soit en s'inspirant de la structure d'un ligand observ exprimentalement.Les modles gnratifs de deep learning ouvrent de nouvelles perspectives pour cette approche centre sur le ligand, car ils permettent d'explorer tout l'espace des molcules synthtisables et de gnrer des composs optimiss.Cependant, ils n'utilisent pas directement la structure de la cible.Nous proposons une mthode pour gnrer des composs de haute affinit une cible, en combinant les avantages des approches par ligands et par cible.OptiMol utilise le docking molculaire comme un oracle, et gnre itrativement des composs prometteurs dans tout l'espace chimique, en amliorant le modle gnratif chaque tape.Ainsi, OptiMol gnre des candidats compatibles avec la structure de la cible, sans s'appuyer sur la connaissance de ligands exprimentaux priori.Pour cela, nous introduisons galement un Auto-encodeur variationnel utilisant la fois les graphes molculaires et les Selfies.Ce modle gnratif gale l'tat de l'art et permet de gnrer des composs 18 fois plus rapidement.Nous montrons qu'OptiMol permet de gnrer une distribution de molcules avec un score de docking amlior, et augmente significativement le nombre de hits trouvs par rapport au docking d'une librairie fixe.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".