MAPS: machine-assisted phenotype scoring enables rapid functional assessment of genetic variants by high-content microscopy
Bibliographic record
Abstract
BACKGROUND: Genetic testing is widely used in evaluating a patient's predisposition to hereditary diseases. In the case of cancer, when a functionally impactful mutation (i.e. genetic variant) is identified in a disease-relevant gene, the patient is at elevated risk of developing a lesion in their lifetime. Unfortunately, as the rate and coverage of genetic testing has accelerated, our ability to assess the functional status of new variants has fallen behind. Therefore, there is an urgent need for more practical, streamlined and cost-effective methods for classifying variants. RESULTS: To directly address this issue, we designed a new approach that uses alterations in protein subcellular localization as a key indicator of loss of function. Thus, new variants can be rapidly functionalized using high-content microscopy (HCM). To facilitate the analysis of the large amounts of imaging data, we developed a new software toolkit, named MAPS for machine-assisted phenotype scoring, that utilizes deep learning to extract and classify cell-level features. MAPS helps users leverage cloud-based deep learning services that are easy to train and deploy to fit their specific experimental conditions. Model training is code-free and can be done with limited training images. Thus, MAPS allows cell biologists to easily incorporate deep learning into their image analysis pipeline. We demonstrated an effective variant functionalization workflow that integrates HCM and MAPS to assess missense variants of PTEN, a tumor suppressor that is frequently mutated in hereditary and somatic cancers. CONCLUSIONS: This paper presents a new way to rapidly assess variant function using cloud deep learning. Since most tumor suppressors have well-defined subcellular localizations, our approach could be widely applied to functionalize variants of uncertain significance and help improve the utility of genetic testing.
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".