Closing the loop: Teaching single-cell foundation models to learn from perturbations
Pershad, Y.; Nandi, T. N.; Van Amburg, J. C.; Parker, A. C.; Ostrowski, L. H.; Giannini, H. K.; Ong, D. A.; Heimlich, J. B.; Obeng, E.; Ericson, K.; Agarwal, A.; Madduri, R. K.; Bick, A. G.
Show abstract
The application of transfer learning models to large scale single-cell datasets has enabled the development of single-cell foundation models (scFMs) that can predict cellular responses to perturbations in silico. Although these predictions can be experimentally tested, current scFMs are unable to "close the loop" and learn from these experiments to create better predictions. Here, we introduce a "closed-loop" framework that extends the scFM by incorporating perturbation data during model fine-tuning. Our closed-loop model improves prediction accuracy, increasing positive predictive value in the setting of T-cell activation three-fold. We applied this model to RUNX1-familial platelet disorder, a rare pediatric blood disorder and identified two therapeutic targets (mTOR and CD74-MIF signaling axis) and two novel pathways (protein kinase C and phosphoinositide 3-kinase). This work establishes that iterative incorporation of experimental data to foundation models enhances biological predictions, representing a crucial step toward realizing the promise of "virtual cell" models for biomedical discovery.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- Sagittarius: Extrapolating Heterogeneous Time-Series Gene Expression Data 95%
- Network-aware self-supervised learning enables high-content phenotypic screening for genetic modifiers of neuronal activity dynamics 94%
- Self-iterative multiple instance learning enables the prediction of CD4+ T cell immunogenic epitopes 93%
Similar papers in this journal
- Integrative, high-resolution analysis of single cell gene expression across experimental conditions with PARAFAC2-RISE 94%
- Conserved epigenetic regulatory logic infers genes governing cell identity 94%
- scTrace+: enhance the cell fate inference by integrating the lineage-tracing and multi-faceted transcriptomic similarity information 93%
Similar papers in this journal
- Combinatorial prediction of therapeutic perturbations using causally-inspired neural networks 93%
- Ultra-fast Prediction of Somatic Structural Variations by Reduced Read Mapping via Pan-Genome k-mer Sets 92%
- Identifying perturbations that boost T-cell infiltration into tumours via counterfactual learning of their spatial proteomic profiles 92%
Similar papers in this journal
- Uncovering the hidden structure of dynamic T cell composition in peripheral blood during cancer immunotherapy: a topic modeling approach 94%
- Parallelized engineering of mutational models using piggyBac transposon delivery of CRISPR libraries 93%
- Quantifying tumor specificity using Bayesian probabilistic modeling for drug target discovery and prioritization 93%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.