A Transformer-Based Deep Learning Approach to Predicting Air Organic Pollutant–Human Protein Interactions
Bibliographic record
Abstract
High Resolution Image Download MS PowerPoint Slide Air pollution poses a critical global public health challenge. Molecular-level initiating events, such as pollutant–protein interactions, can trigger cascades of biological responses that may contribute to adverse health effects. However, current methods are limited in their ability to systematically identify these early binding events, particularly for emerging airborne pollutants, which hinders mechanistic understanding and risk assessment of pollution-related toxicity. To address this, we developed tipFormer (pollutan t –protein i nteraction p rediction based on trans former ), a novel deep learning approach for predicting interactions between airborne organic pollutants and human proteins. The model incorporates dual pretrained language models to encode proteins and organic pollutants, coupled with cross-attention mechanisms to learn intricate interaction patterns underlying pollutant–protein binding. Rigorous validation demonstrated that tipFormer achieves state-of-the-art performance, with an AUC of 0.9787 on a test set. Furthermore, genome-wide transcriptomic analysis using human bronchial epithelial cells exposed to three representative airborne pollutants revealed significant concordance between tipFormer’s predicted targets and the experimentally responsive genes, thereby supporting the model’s biological relevance. By bridging large-scale computational predictions with transcriptomic validation, this study provides deeper mechanistic insight into the molecular basis of air pollution-related adverse outcomes.
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) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 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) | 0.001 | 0.000 |
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".