Biological representation disentanglement of single-cell data
Bibliographic record
Abstract
Abstract Due to its internal state or external environment, a cell’s gene expression profile contains multiple signatures, simultaneously encoding information about its characteristics. Disentangling these factors of variations from single-cell data is needed to recover multiple layers of biological information and extract insight into the individual and collective behavior of cellular populations. While several recent methods were suggested for biological disentanglement, each has its limitations; they are either task-specific, cannot capture inherent nonlinear or interaction effects, cannot integrate layers of experimental data, or do not provide a general reconstruction procedure. We present biolord , a deep generative framework for disentangling known and unknown attributes in single-cell data. Biolord exposes the distinct effects of different biological processes or tissue structure on cellular gene expression. Based on that, biolord allows generating experimentally-inaccessible cell states by virtually shifting cells across time, space, and biological states. Specifically, we showcase accurate predictions of cellular responses to drug perturbations and generalization to predict responses to unseen drugs. Further, biolord disentangles spatial, temporal, and infection-related attributes and their associated gene expression signatures in a single-cell atlas of Plasmodium infection progression in the mouse liver. Biolord can handle partially labeled attributes by predicting a classification for missing labels, and hence can be used to computationally extend an infected hepatocyte population identified at a late stage of the infection to earlier stages. Biolord applies to diverse biological settings, is implemented using the scvi-tools library, and is released as open-source software at https://github.com/nitzanlab/biolord .
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 | 0.001 | 0.002 |
| Research integrity | 0.001 | 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".