Abstract Recent applications of scRNA-seq for massively multiplexed chemical screens have enabled comprehensive profiling of drug responses at unprecedented scale and resolution. However, current assays remain limited to RNA readouts, lacking information on other phenotypic and mechanistic layers such as chromatin accessibility, protein abundance and post-translational modifications. Here, we introduce a scalable framework for multimodal chemical screens, combining parallel small-molecule perturbations with multimodal readouts. We extend existing experimental platforms into icCITE-plex and DOGMA-plex, enabling joint profiling of RNA, protein, and epigenomic responses to thousands of chemical perturbations in parallel. To systematically decode the regulatory circuitry underlying these responses, we develop MoCAVI, a contrastive analysis framework that disentangles the effect of small molecules from control variation in multimodal measurements, and PERCISTRA, a pipeline that infers causal links between chromatin accessibility and gene expression. Applied across ~410,000 primary T cells under ~2,800 conditions, our approach resolves compound-specific mechanisms, highlights off-target effects, and links chromatin accessibility changes to transcription factor networks in primary T cells. Our results establish a generalizable platform for profiling and analyzing cellular responses to chemical perturbations across multiple modalities.

CD4+ T cell responses are exquisitely antigen specific and directed toward peptide epitopes displayed by human leukocyte antigen class II (HLA-II) on antigen-presenting cells. Underrepresentation of diverse alleles in ligand databases and an incomplete understanding of factors affecting antigen presentation in vivo have limited progress in defining principles of peptide immunogenicity. Here, we employed monoallelic immunopeptidomics to identify 358,024 HLA-II binders, with a particular focus on HLA-DQ and HLA-DP. We uncovered peptide-binding patterns across a spectrum of binding affinities and enrichment of structural antigen features. These aspects underpinned the development of context-aware predictor of T cell antigens (CAPTAn), a deep learning model that predicts peptide antigens based on their affinity to HLA-II and full sequence of their source proteins. CAPTAn was instrumental in discovering prevalent T cell epitopes from bacteria in the human microbiome and a pan-variant epitope from SARS-CoV-2. Together CAPTAn and associated datasets present a resource for antigen discovery and the unraveling genetic associations of HLA alleles with immunopathologies.

Plasma cells (PCs) constitute a significant fraction of colonic mucosal cells and contribute to inflammatory infiltrates in ulcerative colitis (UC). While gut PCs secrete bacteria-targeting IgA antibodies, their role in UC pathogenesis is unknown. We performed single-cell V(D)J- and RNA-seq on sorted B cells from the colon of healthy individuals and patients with UC. A large fraction of B cell clones is shared between different colon regions, but inflammation in UC broadly disrupts this landscape, causing transcriptomic changes characterized by an increase in the unfolded protein response (UPR) and antigen presentation genes, clonal expansion, and isotype skewing from IgA1 and IgA2 to IgG1. We also directly expressed and assessed the specificity of 152 mAbs from expanded PC clones. These mAbs show low polyreactivity and autoreactivity and instead target both shared bacterial antigens and specific bacterial strains. Altogether, our results characterize the microbiome-specific colon PC response and how its disruption might contribute to inflammation in UC.

Pooled CRISPR screens with single-cell RNA sequencing readout (Perturb-seq) have emerged as a key technique in functional genomics, but they are limited in scale by cost and combinatorial complexity. In this study, we modified the design of Perturb-seq by incorporating algorithms applied to random, low-dimensional observations. Compressed Perturb-seq measures multiple random perturbations per cell or multiple cells per droplet and computationally decompresses these measurements by leveraging the sparse structure of regulatory circuits. Applied to 598 genes in the immune response to bacterial lipopolysaccharide, compressed Perturb-seq achieves the same accuracy as conventional Perturb-seq with an order of magnitude cost reduction and greater power to learn genetic interactions. We identified known and novel regulators of immune responses and uncovered evolutionarily constrained genes with downstream targets enriched for immune disease heritability, including many missed by existing genome-wide association studies. Our framework enables new scales of interrogation for a foundational method in functional genomics.

E3 ligases regulate key processes, but many of their roles remain unknown. Using Perturb-seq, we interrogated the function of 1,130 E3 ligases, partners and substrates in the inflammatory response in primary dendritic cells (DCs). Dozens impacted the balance of DC1, DC2, migratory DC and macrophage states and a gradient of DC maturation. Family members grouped into co-functional modules that were enriched for physical interactions and impacted specific programs through substrate transcription factors. E3s and their adaptors co-regulated the same processes, but partnered with different substrate recognition adaptors to impact distinct aspects of the DC life cycle. Genetic interactions were more prevalent within than between modules, and a deep learning model, comβVAE, predicts the outcome of new combinations by leveraging modularity. The E3 regulatory network was associated with heritable variation and aberrant gene expression in immune cells in human inflammatory diseases. Our study provides a general approach to dissect gene function.

Pooled CRISPR screens with single-cell RNA-seq readout (Perturb-seq) have emerged as a key technique in functional genomics, but are limited in scale by cost and combinatorial complexity. Here, we reimagine Perturb-seq’s design through the lens of algorithms applied to random, low-dimensional observations. We present compressed Perturb-seq, which measures multiple random perturbations per cell or multiple cells per droplet and computationally decompresses these measurements by leveraging the sparse structure of regulatory circuits. Applied to 598 genes in the immune response to bacterial lipopolysaccharide, compressed Perturb-seq achieves the same accuracy as conventional Perturb-seq at 4 to 20-fold reduced cost, with greater power to learn genetic interactions. We identify known and novel regulators of immune responses and uncover evolutionarily constrained genes with downstream targets enriched for immune disease heritability, including many missed by existing GWAS or trans-eQTL studies. Our framework enables new scales of interrogation for a foundational method in functional genomics.

This study characterizes the plasma cell and monoclonal antibody repertoire throughout the human colon at high resolution using matched single-cell V(D)J- and RNA-seq profiling, as well as antibody characterization of biopsies from patients with ulcerative colitis and healthy controls.

SARS-CoV-2 infection leads to a broad range of outcomes and immune responses, with the development of neutralizing antibodies generally correlated with protection against reinfection. Here, we have characterized both neutralizing activity and T cell responses in a cluster of subjects with mild disease linked to a single spreading event. Surprisingly, we observed sex-specific associations between spike- and particularly nucleoprotein-specific T cell responses and neutralization, with pro-inflammatory cytokines being linked to higher titers only in males. Using single cell immunoprofiling, which provided matched transcriptome and T-cell receptor (TCR) profiles in restimulated CD4 + and CD8 + cells from these subjects, we identified differences in type I IFN signaling that may underlie this difference in antibody generation. Finally, we also identified several TCRs associated with cytokine producing T cells. Altogether, our work maps the breadth of immunological outcomes of SARS-CoV2 infections and highlight the potential role of sex-specific feedback loops during the generation of neutralizing antibodies.

Monoclonal antibodies (mAbs) are a focus in vaccine and therapeutic design to counteract severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its variants. Here, we combined B cell sorting with single-cell VDJ and RNA sequencing (RNA-seq) and mAb structures to characterize B cell responses against SARS-CoV-2. We show that the SARS-CoV-2-specific B cell repertoire consists of transcriptionally distinct B cell populations with cells producing potently neutralizing antibodies (nAbs) localized in two clusters that resemble memory and activated B cells. Cryo-electron microscopy structures of selected nAbs from these two clusters complexed with SARS-CoV-2 spike trimers show recognition of various receptor-binding domain (RBD) epitopes. One of these mAbs, BG10-19, locks the spike trimer in a closed conformation to potently neutralize SARS-CoV-2, the recently arising mutants B.1.1.7 and B.1.351, and SARS-CoV and cross-reacts with heterologous RBDs. Together, our results characterize transcriptional differences among SARS-CoV-2-specific B cells and uncover cross-neutralizing Ab targets that will inform immunogen and therapeutic design against coronaviruses.

Despite their importance in determining protein abundance, a comprehensive catalogue of sequence features controlling protein-to-mRNA (PTR) ratios and a quantification of their effects are still lacking. Here, we quantified PTR ratios for 11,575 proteins across 29 human tissues using matched transcriptomes and proteomes. We estimated by regression the contribution of known sequence determinants of protein synthesis and degradation in addition to 45 mRNA and 3 protein sequence motifs that we found by association testing. While PTR ratios span more than 2 orders of magnitude, our integrative model predicts PTR ratios at a median precision of 3.2-fold. A reporter assay provided functional support for two novel UTR motifs, and an immobilized mRNA affinity competition-binding assay identified motif-specific bound proteins for one motif. Moreover, our integrative model led to a new metric of codon optimality that captures the effects of codon frequency on protein synthesis and degradation. Altogether, this study shows that a large fraction of PTR ratio variation in human tissues can be predicted from sequence, and it identifies many new candidate post-transcriptional regulatory elements.

Genome-, transcriptome- and proteome-wide measurements provide insights into how biological systems are regulated. However, fundamental aspects relating to which human proteins exist, where they are expressed and in which quantities are not fully understood. Therefore, we generated a quantitative proteome and transcriptome abundance atlas of 29 paired healthy human tissues from the Human Protein Atlas project representing human genes by 18,072 transcripts and 13,640 proteins including 37 without prior protein-level evidence. The analysis revealed that hundreds of proteins, particularly in testis, could not be detected even for highly expressed mRNAs, that few proteins show tissue-specific expression, that strong differences between mRNA and protein quantities within and across tissues exist and that protein expression is often more stable across tissues than that of transcripts. Only 238 of 9,848 amino acid variants found by exome sequencing could be confidently detected at the protein level showing that proteogenomics remains challenging, needs better computational methods and requires rigorous validation. Many uses of this resource can be envisaged including the study of gene/protein expression regulation and biomarker specificity evaluation.

Opportunistic scheduling algorithms in cognitive radio networks (CRNs) allocate resources by exploiting the variations in channel conditions and spectral activities of primary users. However, most of these scheduling algorithms ignore the per-user throughput requirements. In this paper, we formulate a scheduling problem called maximizing the number of satisfied users (MNSU), which maximizes the number of secondary users that are satisfied in terms of throughput in a centralized CRN. We show that MNSU is NP-hard in the strong sense and cannot be approximated within any constant factor better than 2 unless P = NP . We also prove that MNSU is at least as hard as the max-min fair scheduling problem, which has previously been proven to be a computationally very difficult problem in the literature. We then propose two heuristic algorithms: 1) best first resource assignment and 2) resource assignment with partial backtracking. We demonstrate that our proposed algorithms yield high performance while still achieving low computational complexity.

In this letter we propose an auction theory based algorithm for throughput maximizing scheduling in centralized cognitive radio networks (CRN). In the considered CRN scheme, a centralized base station coordinates the assignment of frequencies and time slots to cognitive users with multiple antennas. Our proposed algorithm uses first-price sealed bid auction mechanism in which frequency and time slot pairs are considered as the auctioned resources and cognitive users are the bidders. The experimental results show that our computationally efficient algorithm yields very close throughput performance to the optimization software CPLEX values.