Benjamin Kuznets-Speck

ExPert infers causal perturbations from cell states by learning prototypes across 12 contexts, achieving high recall with low compute and validating findings in disease models.

This paper introduces TwINFER, a framework using twin cell information to improve gene regulatory network inference, resolving directionality and reducing false positives in network motifs.

Keeping SCORE transforms diffusion models into probabilistic classifiers for genomics, providing accurate predictions with uncertainty and interpretability across multiple tasks.

CIPHER uses gene co-fluctuations in unperturbed cells to predict transcriptome-wide responses to perturbations, outperforming existing methods and revealing key regulatory structures.

The paper develops TWAVE, a generative model that identifies causal gene sets for complex traits by integrating transcriptomic data and gene perturbation responses.

The paper analyzes the energetic costs of cellular signaling, showing that ATP consumption and chemical potential determine the information capacity of kinase-phosphatase loops, with yeast signaling pathways approaching optimality.