Rademaker, T. J. | bioRxiv Science

bioRxiv2021

Pre-existing chromatin accessibility and gene expression differences among naïve CD4+ T cells influence effector potential

CD4+ T cells have a remarkable potential to differentiate into diverse effector lineages following activation. Here, we probed the heterogeneity present among naive CD4+ T cells before encountering their cognate antigen to ask whether their effector potential is modulated by pre-existing transcriptional and epigenetic differences. Using single-cell RNA sequencing, we showed that key drivers of variability are genes involved in T cell receptor (TCR) signaling. Using CD5 expression as a read-out of the strength of tonic TCR interactions with self-peptide MHC, and sorting on the ends of this self-reactivity spectrum, we find that pre-existing transcriptional differences among naive CD4+ T cells impact follicular helper cell (TFH) versus non-TFH effector lineage choice. Moreover, our data implicate TCR signal strength during thymic development in establishing differences in naive CD4 T cell chromatin landscapes that ultimately shape their effector potential.

immunology↗

bioRxiv2018

Fooling the classifier: Ligand antagonism and adversarial examples

Machine learning algorithms can be fooled by small well-designed adversarial perturbations. This is reminiscent of cellular decision-making where ligands (called antagonists) prevent correct signalling, like in early immune recognition. We draw a formal analogy between neural networks used in machine learning and models of cellular decision-making (adaptive proofreading). We apply attacks from machine learning to simple decision-making models, and show explicitly the correspondence to antagonism by weakly bound ligands. Such antagonism is absent in more nonlinear models, which inspired us to implement a biomimetic defence in neural networks filtering out adversarial perturbations. We then apply a gradient-descent approach from machine learning to different cellular decision-making models, and we reveal the existence of two regimes characterized by the presence or absence of a critical point for the gradient. This critical point causes the strongest antagonists to lie close to the decision boundary. This is validated in the loss landscapes of robust neural networks and cellular decision-making models, and observed experimentally for immune cells. For both regimes, we explain how associated defence mechanisms shape the geometry of the loss landscape, and why different adversarial attacks are effective in different regimes. Our work connects evolved cellular decision-making to machine learning, and motivates the design of a general theory of adversarial perturbations, both for in vivo and in silico systems.

biophysics↗