Major cell-types in multiomic single-nucleus datasets impact statistical modeling of links between regulatory sequences and target genes
Bibliographic record
Abstract
Epigenomic profiling, including ATACseq, is one of the main tools used to define enhancers. Because enhancers are overwhelmingly cell-type specific, inference of their activity is greatly limited in complex tissues. Multiomic assays that probe in the same nucleus both the open chromatin landscape and gene expression levels enable the study of correlations (links) between these two modalities. Current best practices to infer the regulatory effect of candidate cis-regulatory elements (cCREs) in multiomic data involve removing biases associated with GC content by generating null distributions of matched ATACseq peaks drawn from different chromosomes. This strategy has been broadly adopted by popular single-nucleus multiomic workflows such as Signac. Here, we uncovered limitations and confounders of this approach. We found a strong loss of power to detect a regulatory effect for cCREs with high read counts in the dominant cell-type. We showed that this is largely due to cell-type-specific trans-ATACseq peak correlations creating bimodal null distributions. We tested alternative models and concluded that physical distance and/or the raw Pearson correlation coefficients are the best predictors for peak-gene links when compared to predictions from Epimap (e.g. CD14 area under the curve [AUC] = 0.51 with the method implemented in Signac vs. 0.71 with the Pearson correlation coefficients) or validation by CRISPR perturbations (AUC = 0.63 vs. 0.73).
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 | 0.010 | 0.021 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| 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".