A Premalignant Cell-Based Model for Functionalization and Classification of <i>PTEN</i> Variants
Bibliographic record
Abstract
Abstract As sequencing becomes more economical, we are identifying sequence variations in the population faster than ever. For disease-associated genes, it is imperative that we differentiate a sequence variant as either benign or pathogenic, such that the appropriate therapeutic interventions or surveillance can be implemented. PTEN is a frequently mutated tumor suppressor that has been linked to the PTEN hamartoma tumor syndrome. Although the domain structure of PTEN and the functional impact of a number of its most common tumor-linked mutations have been characterized, there is a lack of information about many recently identified clinical variants. To address this challenge, we developed a cell-based assay that utilizes a premalignant phenotype of normal mammary epithelial cells lacking PTEN. We measured the ability of PTEN variants to rescue the spheroid formation phenotype of PTEN−/− MCF10A cells maintained in suspension. As proof of concept, we functionalized 47 missense variants using this assay, only 19 of which have clear classifications in ClinVar. We utilized a machine learning model trained with annotated genotypic data to classify variants as benign or pathogenic based on our functional scores. Our model predicted with high accuracy that loss of PTEN function was indicative of pathogenicity. We also determined that the pathogenicity of certain variants may have arisen from reduced stability of the protein product. Overall, this assay outperformed computational predictions, was scalable, and had a short run time, serving as an ideal alternative for annotating the clinical significance of cancer-associated PTEN variants. Significance: Combined three-dimensional tumor spheroid modeling and machine learning classifies PTEN missense variants, over 70% of which are currently listed as variants of uncertain significance.
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 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| 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".