Armita Nourmohammad

Bio: Armita Nourmohammad is an academic researcher from University of Washington. The author has contributed to research in topics: Acquired immune system & Artificial neural network. The author has an hindex of 5, co-authored 18 publications receiving 172 citations. Previous affiliations of Armita Nourmohammad include Fred Hutchinson Cancer Research Center & Max Planck Society.

Gene expression variability across cells and species shapes innate immunity

Abstract: As the first line of defence against pathogens, cells mount an innate immune response, which varies widely from cell to cell. The response must be potent but carefully controlled to avoid self-damage. How these constraints have shaped the evolution of innate immunity remains poorly understood. Here we characterize the innate immune response’s transcriptional divergence between species and variability in expression among cells. Using bulk and single-cell transcriptomics in fibroblasts and mononuclear phagocytes from different species, challenged with immune stimuli, we map the architecture of the innate immune response. Transcriptionally diverging genes, including those that encode cytokines and chemokines, vary across cells and have distinct promoter structures. Conversely, genes that are involved in the regulation of this response, such as those that encode transcription factors and kinases, are conserved between species and display low cell-to-cell variability in expression. We suggest that this expression pattern, which is observed across species and conditions, has evolved as a mechanism for fine-tuned regulation to achieve an effective but balanced response.

Major antigenic site B of human influenza H3N2 viruses has an evolving local fitness landscape.

Abstract: Antigenic drift of influenza virus hemagglutinin (HA) is enabled by facile evolvability. However, HA antigenic site B, which has become immunodominant in recent human H3N2 influenza viruses, is also evolutionarily constrained by its involvement in receptor binding. Here, we employ deep mutational scanning to probe the local fitness landscape of HA antigenic site B in six different human H3N2 strains spanning from 1968 to 2016. We observe that the fitness landscape of HA antigenic site B can be very different between strains. Sequence variants that exhibit high fitness in one strain can be deleterious in another, indicating that the evolutionary constraints of antigenic site B have changed over time. Structural analysis suggests that the local fitness landscape of antigenic site B can be reshaped by natural mutations via modulation of the receptor-binding mode. Overall, these findings elucidate how influenza virus continues to explore new antigenic space despite strong functional constraints. Antigenic site B in influenza A virus hemagglutinin (HA) is immunodominant in circulating human H3N2 strains. Using deep mutational scanning, Wu et al. here define the local fitness landscapes of HA antigenic site B in six human H3N2 strains, providing insights into evolvability of influenza antigenicity.

Optimal Evolutionary Control for Artificial Selection on Molecular Phenotypes

Abstract: Optimal control for artificial selection offers a new paradigm for directing stochastic evolution of multivariate molecular characteristics and phenotypes toward desired targets.

Generative models of T-cell receptor sequences

Abstract: T-cell receptors (TCR) are key proteins of the adaptive immune system, generated randomly in each individual, whose diversity underlies our ability to recognize infections and malignancies. Modeling the distribution of TCR sequences is of key importance for immunology and medical applications. Here, we compare two inference methods trained on high-throughput sequencing data: a knowledge-guided approach, which accounts for the details of sequence generation, supplemented by a physics-inspired model of selection; and a knowledge-free variational autoencoder based on deep artificial neural networks. We show that the knowledge-guided model outperforms the deep network approach at predicting TCR probabilities, while being more interpretable, at a lower computational cost.

A High-Resolution Map of Human Evolutionary Constraint Using 29 Mammals

Abstract: The comparison of related genomes has emerged as a powerful lens for genome interpretation. Here we report the sequencing and comparative analysis of 29 eutherian genomes. We confirm that at least 5.5% of the human genome has undergone purifying selection, and locate constrained elements covering ∼4.2% of the genome. We use evolutionary signatures and comparisons with experimental data sets to suggest candidate functions for ∼60% of constrained bases. These elements reveal a small number of new coding exons, candidate stop codon readthrough events and over 10,000 regions of overlapping synonymous constraint within protein-coding exons. We find 220 candidate RNA structural families, and nearly a million elements overlapping potential promoter, enhancer and insulator regions. We report specific amino acid residues that have undergone positive selection, 280,000 non-coding elements exapted from mobile elements and more than 1,000 primate- and human-accelerated elements. Overlap with disease-associated variants indicates that our findings will be relevant for studies of human biology, health and disease.

scGen predicts single-cell perturbation responses

Abstract: Accurately modeling cellular response to perturbations is a central goal of computational biology. While such modeling has been based on statistical, mechanistic and machine learning models in specific settings, no generalization of predictions to phenomena absent from training data (out-of-sample) has yet been demonstrated. Here, we present scGen (https://github.com/theislab/scgen), a model combining variational autoencoders and latent space vector arithmetics for high-dimensional single-cell gene expression data. We show that scGen accurately models perturbation and infection response of cells across cell types, studies and species. In particular, we demonstrate that scGen learns cell-type and species-specific responses implying that it captures features that distinguish responding from non-responding genes and cells. With the upcoming availability of large-scale atlases of organs in a healthy state, we envision scGen to become a tool for experimental design through in silico screening of perturbation response in the context of disease and drug treatment.

A temporal switch in the germinal center determines differential output of memory B and plasma cells (IRC3P.468)

Abstract: Though memory B cells (MBCs) and long-lived plasma cells (LLPCs) are both thought to derive from the germinal center (GC) reaction, there is little insight into or agreement about the signals that control differentiation to one cell type or another. By performing BrdU pulse-labeling studies, GC disruption experiments and V gene sequencing, we found that the generation of these cell types is actually temporally controlled and separated during the immune response. We report that MBCs mainly derive from early GCs (much before GC peak size), while more affinity matured LLPCs are predominantly formed during late GCs - long after its peak in size. Based on these findings, we propose a new model that the GC response undergoes a temporal switch, functioning quite differently at early and late stages. Therefore the generation of MBCs and LLPCs is the consequence of a general shift in GC output over time rather than the result of specific instructive signals that are selectively delivered to GC B cells at any given time during the response. We also present direct evidence that a large fraction of long-lived IgM+ MBC, and even some IgG+ MBC, are formed at very early time points, prior to the existence of detectable GCs. The knowledge of when specific long-lived immune-effector cells are generated during an immune response has strong implications for vaccine design and understanding long-term pathogen immunity.

Confronting false discoveries in single-cell differential expression.

Abstract: Differential expression analysis in single-cell transcriptomics enables the dissection of cell-type-specific responses to perturbations such as disease, trauma, or experimental manipulations. While many statistical methods are available to identify differentially expressed genes, the principles that distinguish these methods and their performance remain unclear. Here, we show that the relative performance of these methods is contingent on their ability to account for variation between biological replicates. Methods that ignore this inevitable variation are biased and prone to false discoveries. Indeed, the most widely used methods can discover hundreds of differentially expressed genes in the absence of biological differences. To exemplify these principles, we exposed true and false discoveries of differentially expressed genes in the injured mouse spinal cord.

Extreme heterogeneity of influenza virus infection in single cells

Abstract: When viruses infect cells, they take over the cell’s machinery and use it to express their own genes. This process has mostly been studied by looking at the average outcome of infection when many viruses infect many cells. However, it is less clear what happens in individual cells. For example, does the virus take over every cell to make lots of viral gene products, or do some cells produce far more viral gene products than others? Russell et al. have now used a new technique called single-cell RNA sequencing to look at how well influenza virus genes were expressed in hundreds of individual mammalian cells. The goal was to work out how the outcome of infection varied between different cells. One way to quantify variability – also known as heterogeneity – is by using a statistical measure called the Gini coefficient. This statistic is often used to assess the inequality in incomes across a nation.In the hypothetical situation where everyone earned the same income, the Gini coefficient would equal zero; while if only one person had all the income and all others had none, the value would be very close to one. In reality, countries fall somewhere in between these two extremes. In the United States for instance, the Gini coefficient for income is 0.47. When Russell et al. worked out the Gini coefficient for the amount of viral genes expressed in different cells, the value was at least 0.64. This indicates that there is more unevenness in viral gene expression for influenza than there is income inequality in the United States. So, what characterizes the “Bill Gates” cells and viruses that have the highest viral gene expression? Influenza viruses sometimes fail to express some of their genes. Russell et al. found that this failure often led to “poor” viruses that were less productive than “rich” viruses that expressed all the critical genes. However, the results suggest that there are also other factors that contribute a lot to the heterogeneity. Real influenza virus infections are usually started by very few viruses, so this new understanding of the variability that occurs when individual viruses infect individual cells might prove important for understanding the properties of infections at larger scales too.