LSTM-driven drug design using SELFIES for target-focused de novo generation of HIV-1 protease inhibitor candidates for AIDS treatment
Bibliographic record
Abstract
The battle against viral drug resistance highlights the need for innovative approaches to replace time-consuming and costly traditional methods. Deep generative models offer automation potential, especially in the fight against Human immunodeficiency virus (HIV), as they can synthesize diverse molecules effectively. In this paper, an application of an LSTM-based deep generative model named "LSTM-ProGen" is proposed to be tailored explicitly for the de novo design of drug candidate molecules that interact with a specific target protein (HIV-1 protease). LSTM-ProGen distinguishes itself by employing a long-short-term memory (LSTM) architecture, to generate novel molecules target specificity against the HIV-1 protease. Following a thorough training process involves fine-tuning LSTM-ProGen on a diverse range of compounds sourced from the ChEMBL database. The model was optimized to meet specific requirements, with multiple iterations to enhance its predictive capabilities and ensure it generates molecules that exhibit favorable target interactions. The training process encompasses an array of performance evaluation metrics, such as drug-likeness properties. Our evaluation includes extensive silico analysis using molecular docking and PCA-based visualization to explore the chemical space that the new molecules cover compared to those in the training set. These evaluations reveal that a subset of 12 de novo molecules generated by LSTM-ProGen exhibit a striking ability to interact with the target protein, rivaling or even surpassing the efficacy of native ligands. Extended versions with further refinement of LSTM-ProGen hold promise as versatile tools for designing efficacious and customized drug candidates tailored to specific targets, thus accelerating drug development and facilitating the discovery of new therapies for various diseases.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".