Improving Molecular De Novo Drug Design with Transformers
Bibliographic record
Abstract
Drug design is undergoing a transformation as we challenge conventional methods by integrating state-of-the-art artificial intelligence with the intricate domain of molecular biology. At the heart of our endeavor lies a significant challenge: the scarcity of datasets containing active compounds for emerging target proteins. To confront this obstacle, we're pioneering an innovative approach. We're merging the advanced Generative Pre-trained Transformer (GPT) architecture with the nuanced capabilities of Long Short-Term Memory (LSTM) networks, with the aim of generating Simplified Molecular Input Line Entry System (SMILES) strings to unveil novel therapeutic pathways. Additionally, we're employing a Bidirectional Encoder Representations from Transformers (BERT) pretraining strategy to enrich our model with comprehensive molecular data, including amino acid sequences and molecular SMILES datasets. Through meticulous fine-tuning on a meticulously curated protein-ligand complex dataset, we're achieving precise conditional generation via autoregressive supervised learning. Our research introduces a groundbreaking method to assess molecular affinity, validated against established proteins, showcasing superior binding affinities compared to certain FDA-approved drugs in docking experiments. By pushing the boundaries of generative algorithms and establishing a robust framework for evaluating molecular affinity, we're driving forward the field of de novo drug design, offering promising therapeutic avenues and enabling deeper exploration of the chemical landscape.
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 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 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".