protPheMut: An Interpretable Machine Learning Tool for Classification of Cancer and Neurodevelopmental Disorders in Human Missense Mutations
Bibliographic record
Abstract
Recent advances in human genomics have revealed that missense mutations in a single protein can lead to distinctly different phenotypes. In particular, some mutations in oncoproteins like MEK1, MEK2, PI3Kα, PTEN, SHAP2, and RAS are linked various cancers and neurodevelopmental disorders (NDDs). While numerous tools exist for predicting the pathogenicity of missense mutations, linking these mutations to certain phenotypes remains a major challenge, particularly in the context of personalized medicine. To fill this gap, we developed protPheMut (Protein Phenotypic Mutations Analyzer, http://netprotlab.com/protPheMut), leveraging interpretable machine learning approaches and enhancing model transparency through SHAP explanations, to integrate diverse biophysical and network dynamics-based signatures for predicting whether mutations in the same protein promote cancer or NDDs. Overall, proPheMut achieved an AUCROC of 0.9118 in cross-validation and 0.8925 on an independent test set for discriminating cancer- versus NDDs-related mutations. We further illustrate its utility in phenotype (cancer/NDDs) prediction by mutation analyses of two protein cases, PI3Kα and PTEN. Compared to seven other predictive tools, protPheMut demonstrated exceptional accuracy in forecasting phenotypic effects, achieving an AUROC of 0.8501 for PI3Kα mutations related to cancer and Cowden syndrome. For multi-phenotype prediction of PTEN mutations related to cancer, PHTS, and HCPS, protPheMut achieved an AUROC of 0.9349 through micro averaging. Using SHAP model explanations, protPheMut highlights the strength of network and dynamic features in deeper uncovering of the effects of pathogenic mutations, thus classifying different disease phenotypes.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".