Computational prediction of protein interactions in single cells by proximity sequencing.

Proximity sequencing (Prox-seq) simultaneously measures gene expression, protein expression and protein complexes on single cells. Using information from dual-antibody binding events, Prox-seq infers surface protein dimers at the single-cell level. Prox-seq provides multi-dimensional phenotyping of single cells in high throughput, and was recently used to track the formation of receptor complexes during cell signaling and discovered a novel interaction between CD9 and CD8 in naïve T cells. The distribution of protein abundance can affect identification of protein complexes in a complicated manner in dual-binding assays like Prox-seq. These effects are difficult to explore with experiments, yet important for accurate quantification of protein complexes. Here, we introduce a physical model of Prox-seq and computationally evaluate several different methods for reducing background noise when quantifying protein complexes. Furthermore, we developed an improved method for analysis of Prox-seq data, which resulted in more accurate and robust quantification of protein complexes. Finally, our Prox-seq model offers a simple way to investigate the behavior of Prox-seq data under various biological conditions and guide users toward selecting the best analysis method for their data.

TRIBAL: Tree Inference of B cell Clonal Lineages.

B cells are a critical component of the adaptive immune system, responsible for producing antibodies that help protect the body from infections and foreign substances. Single cell RNA-sequencing (scRNA-seq) has allowed for both profiling of B cell receptor (BCR) sequences and gene expression. However, understanding the adaptive and evolutionary mechanisms of B cells in response to specific stimuli remains a significant challenge in the field of immunology. We introduce a new method, TRIBAL, which aims to infer the evolutionary history of clonally related B cells from scRNA-seq data. The key insight of TRIBAL is that inclusion of isotype data into the B cell lineage inference problem is valuable for reducing phylogenetic uncertainty that arises when only considering the receptor sequences. Consequently, the TRIBAL inferred B cell lineage trees jointly capture the somatic mutations introduced to the B cell receptor during affinity maturation and isotype transitions during class switch recombination. In addition, TRIBAL infers isotype transition probabilities that are valuable for gaining insight into the dynamics of class switching. Via in silico experiments, we demonstrate that TRIBAL infers isotype transition probabilities with the ability to distinguish between direct versus sequential switching in a B cell population. This results in more accurate B cell lineage trees and corresponding ancestral sequence and class switch reconstruction compared to competing methods. Using real-world scRNA-seq datasets, we show that TRIBAL recapitulates expected biological trends in a model affinity maturation system. Furthermore, the B cell lineage trees inferred by TRIBAL were equally plausible for the BCR sequences as those inferred by competing methods but yielded lower entropic partitions for the isotypes of the sequenced B cell. Thus, our method holds the potential to further advance our understanding of vaccine responses, disease progression, and the identification of therapeutic antibodies.

ATAT: Automated Tissue Alignment and Traversal in Spatial Transcriptomics with Self-Supervised Learning.

Spatial transcriptomics (ST) has enhanced RNA analysis in tissue biopsies, but interpreting these data is challenging without expert input. We present Automated Tissue Alignment and Traversal (ATAT), a novel computational framework designed to enhance ST analysis in the context of multiple and complex tissue architectures and morphologies, such as those found in biopsies of the gastrointestinal tract. ATAT utilizes self-supervised contrastive learning on hematoxylin and eosin (H&E) stained images to automate the alignment and traversal of ST data. This approach addresses a critical gap in current ST analysis methodologies, which rely heavily on manual annotation and pathologist expertise to delineate regions of interest for accurate gene expression modeling. Our framework not only streamlines the alignment of multiple ST samples, but also demonstrates robustness in modeling gene expression transitions across specific regions. Additionally, we highlight the ability of ATAT to traverse complex tissue topologies in real-world cases from various individuals and conditions. Our method successfully elucidates differences in immune infiltration patterns across the intestinal wall, enabling the modeling of transcriptional changes across histological layers. We show that ATAT achieves comparable performance to the state-of-the-art method, while alleviating the burden of manual annotation and enabling alignment of tissue samples with complex morphologies.

Computational prediction of protein interactions on single cells by proximity sequencing.

Proximity sequencing (Prox-seq) measures gene expression, protein expression, and protein complexes at the single cell level, using information from dual-antibody binding events and a single cell sequencing readout. Prox-seq provides multi-dimensional phenotyping of single cells and was recently used to track the formation of receptor complexes during inflammatory signaling in macrophages and to discover a new interaction between CD9/CD8 proteins on naïve T cells. The distribution of protein abundance affects identification of protein complexes in a complicated manner in dual-binding assays like Prox-seq. These effects are difficult to explore with experiments, yet important for accurate quantification of protein complexes. Here, we introduce a physical model for protein dimer formation on single cells and computationally evaluate several different methods for reducing background noise when quantifying protein complexes. Furthermore, we developed an improved method for analysis of Prox-seq single-cell data, which resulted in more accurate and robust quantification of protein complexes. Finally, our model offers a simple way to investigate the behavior of Prox-seq under various biological conditions and guide users toward selecting the best analysis method for their data.

Deploying synthetic coevolution and machine learning to engineer protein-protein interactions.

Fine-tuning of protein-protein interactions occurs naturally through coevolution, but this process is difficult to recapitulate in the laboratory. We describe a platform for synthetic protein-protein coevolution that can isolate matched pairs of interacting muteins from complex libraries. This large dataset of coevolved complexes drove a systems-level analysis of molecular recognition between Z domain-affibody pairs spanning a wide range of structures, affinities, cross-reactivities, and orthogonalities, and captured a broad spectrum of coevolutionary networks. Furthermore, we harnessed pretrained protein language models to expand, in silico, the amino acid diversity of our coevolution screen, predicting remodeled interfaces beyond the reach of the experimental library. The integration of these approaches provides a means of simulating protein coevolution and generating protein complexes with diverse molecular recognition properties for biotechnology and synthetic biology.

Microbiota-dependent proteolysis of gluten subverts diet-mediated protection against type 1 diabetes.

Diet and commensals can affect the development of autoimmune diseases like type 1 diabetes (T1D). However, whether dietary interventions are microbe-mediated was unclear. We found that a diet based on hydrolyzed casein (HC) as a protein source protects non-obese diabetic (NOD) mice in conventional and germ-free (GF) conditions via improvement in the physiology of insulin-producing cells to reduce autoimmune activation. The addition of gluten (a cereal protein complex associated with celiac disease) facilitates autoimmunity dependent on microbial proteolysis of gluten: T1D develops in GF animals monocolonized with Enterococcus faecalis harboring secreted gluten-digesting proteases but not in mice colonized with protease deficient bacteria. Gluten digestion by E. faecalis generates T cell-activating peptides and promotes innate immunity by enhancing macrophage reactivity to lipopolysaccharide (LPS). Gnotobiotic NOD Toll4-negative mice monocolonized with E. faecalis on an HC + gluten diet are resistant to T1D. These findings provide insights into strategies to develop dietary interventions to help protect humans against autoimmunity.

Gut commensal bacteria enhance pathogenesis of a tumorigenic murine retrovirus.

The influence of the microbiota on viral transmission and replication is well appreciated. However, its impact on retroviral pathogenesis outside of transmission/replication control remains unknown. Using murine leukemia virus (MuLV), we found that some commensal bacteria promoted the development of leukemia induced by this retrovirus. The promotion of leukemia development by commensals is due to suppression of the adaptive immune response through upregulation of several negative regulators of immunity. These negative regulators include Serpinb9b and Rnf128, which are associated with a poor prognosis of some spontaneous human cancers. Upregulation of Serpinb9b is mediated by sensing of bacteria by the NOD1/NOD2/RIPK2 pathway. This work describes a mechanism by which the microbiota enhances tumorigenesis within gut-distant organs and points at potential targets for cancer therapy.

Integration of tumor extrinsic and intrinsic features associates with immunotherapy response in non-small cell lung cancer.

The efficacy of immune checkpoint blockade (ICB) varies greatly among metastatic non-small cell lung cancer (NSCLC) patients. Loss of heterozygosity at the HLA-I locus (HLA-LOH) has been identified as an important immune escape mechanism. However, despite HLA-I disruptions in their tumor, many patients have durable ICB responses. Here we seek to identify HLA-I-independent features associated with ICB response in NSCLC. We use single-cell profiling to identify tumor-infiltrating, clonally expanded CD4+ T cells that express a canonical cytotoxic gene program and NSCLC cells with elevated HLA-II expression. We postulate cytotoxic CD4+ T cells mediate anti-tumor activity via HLA-II on tumor cells and augment HLA-I-dependent cytotoxic CD8+ T cell interactions to drive ICB response in NSCLC. We show that integrating tumor extrinsic cytotoxic gene expression with tumor mutational burden is associated with longer time to progression in a real-world cohort of 123 NSCLC patients treated with ICB regimens, including those with HLA-LOH.

Privacy preserving validation for multiomic prediction models.

Reproducibility of results obtained using ribonucleic acid (RNA) data across labs remains a major hurdle in cancer research. Often, molecular predictors trained on one dataset cannot be applied to another due to differences in RNA library preparation and quantification, which inhibits the validation of predictors across labs. While current RNA correction algorithms reduce these differences, they require simultaneous access to patient-level data from all datasets, which necessitates the sharing of training data for predictors when sharing predictors. Here, we describe SpinAdapt, an unsupervised RNA correction algorithm that enables the transfer of molecular models without requiring access to patient-level data. It computes data corrections only via aggregate statistics of each dataset, thereby maintaining patient data privacy. Despite an inherent trade-off between privacy and performance, SpinAdapt outperforms current correction methods, like Seurat and ComBat, on publicly available cancer studies, including TCGA and ICGC. Furthermore, SpinAdapt can correct new samples, thereby enabling unbiased evaluation on validation cohorts. We expect this novel correction paradigm to enhance research reproducibility and to preserve patient privacy.

Innate-like self-reactive B cells infiltrate human renal allografts during transplant rejection.

Intrarenal B cells in human renal allografts indicate transplant recipients with a poor prognosis, but how these cells contribute to rejection is unclear. Here we show using single-cell RNA sequencing that intrarenal class-switched B cells have an innate cell transcriptional state resembling mouse peritoneal B1 or B-innate (Bin) cells. Antibodies generated by Bin cells do not bind donor-specific antigens nor are they enriched for reactivity to ubiquitously expressed self-antigens. Rather, Bin cells frequently express antibodies reactive with either renal-specific or inflammation-associated antigens. Furthermore, local antigens can drive Bin cell proliferation and differentiation into plasma cells expressing self-reactive antibodies. These data show a mechanism of human inflammation in which a breach in organ-restricted tolerance by infiltrating innate-like B cells drives local tissue destruction.

Validation of a liquid biopsy assay with molecular and clinical profiling of circulating tumor DNA.

Liquid biopsy is a valuable precision oncology tool that is increasingly used as a non-invasive approach to identify biomarkers, detect resistance mutations, monitor disease burden, and identify early recurrence. The Tempus xF liquid biopsy assay is a 105-gene, hybrid-capture, next-generation sequencing (NGS) assay that detects single-nucleotide variants, insertions/deletions, copy number variants, and chromosomal rearrangements. Here, we present extensive validation studies of the xF assay using reference standards, cell lines, and patient samples that establish high sensitivity, specificity, and accuracy in variant detection. The Tempus xF assay is highly concordant with orthogonal methods, including ddPCR, tumor tissue-based NGS assays, and another commercial plasma-based NGS assay. Using matched samples, we developed a dynamic filtering method to account for germline mutations and clonal hematopoiesis, while significantly decreasing the number of false-positive variants reported. Additionally, we calculated accurate circulating tumor fraction estimates (ctFEs) using the Off-Target Tumor Estimation Routine (OTTER) algorithm for targeted-panel sequencing. In a cohort of 1,000 randomly selected cancer patients who underwent xF testing, we found that ctFEs correlated with disease burden and clinical outcomes. These results highlight the potential of serial testing to monitor treatment efficacy and disease course, providing strong support for incorporating liquid biopsy in the management of patients with advanced disease.

A pan-cancer organoid platform for precision medicine.

Patient-derived tumor organoids (TOs) are emerging as high-fidelity models to study cancer biology and develop novel precision medicine therapeutics. However, utilizing TOs for systems-biology-based approaches has been limited by a lack of scalable and reproducible methods to develop and profile these models. We describe a robust pan-cancer TO platform with chemically defined media optimized on cultures acquired from over 1,000 patients. Crucially, we demonstrate tumor genetic and transcriptomic concordance utilizing this approach and further optimize defined minimal media for organoid initiation and propagation. Additionally, we demonstrate a neural-network-based high-throughput approach for label-free, light-microscopy-based drug assays capable of predicting patient-specific heterogeneity in drug responses with applicability across solid cancers. The pan-cancer platform, molecular data, and neural-network-based drug assay serve as resources to accelerate the broad implementation of organoid models in precision medicine research and personalized therapeutic profiling programs.

Pseudocell Tracer-A method for inferring dynamic trajectories using scRNAseq and its application to B cells undergoing immunoglobulin class switch recombination.

Single cell RNA sequencing (scRNAseq) can be used to infer a temporal ordering of cellular states. Current methods for the inference of cellular trajectories rely on unbiased dimensionality reduction techniques. However, such biologically agnostic ordering can prove difficult for modeling complex developmental or differentiation processes. The cellular heterogeneity of dynamic biological compartments can result in sparse sampling of key intermediate cell states. To overcome these limitations, we develop a supervised machine learning framework, called Pseudocell Tracer, which infers trajectories in pseudospace rather than in pseudotime. The method uses a supervised encoder, trained with adjacent biological information, to project scRNAseq data into a low-dimensional manifold that maps the transcriptional states a cell can occupy. Then a generative adversarial network (GAN) is used to simulate pesudocells at regular intervals along a virtual cell-state axis. We demonstrate the utility of Pseudocell Tracer by modeling B cells undergoing immunoglobulin class switch recombination (CSR) during a prototypic antigen-induced antibody response. Our results revealed an ordering of key transcription factors regulating CSR to the IgG1 isotype, including the concomitant expression of Nfkb1 and Stat6 prior to the upregulation of Bach2 expression. Furthermore, the expression dynamics of genes encoding cytokine receptors suggest a poised IL-4 signaling state that preceeds CSR to the IgG1 isotype.

Improved integration of single-cell transcriptome and surface protein expression by LinQ-View.

Multimodal advances in single-cell sequencing have enabled the simultaneous quantification of cell surface protein expression alongside unbiased transcriptional profiling. Here, we present LinQ-View, a toolkit designed for multimodal single-cell data visualization and analysis. LinQ-View integrates transcriptional and cell surface protein expression profiling data to reveal more accurate cell heterogeneity and proposes a quantitative metric for cluster purity assessment. Through comparison with existing multimodal methods on multiple public CITE-seq datasets, we demonstrate that LinQ-View efficiently generates accurate cell clusters, especially in CITE-seq data with routine numbers of surface protein features, by preventing variations in a single surface protein feature from affecting results. Finally, we utilized this method to integrate single-cell transcriptional and protein expression data from SARS-CoV-2-infected patients, revealing antigen-specific B cell subsets after infection. Our results suggest LinQ-View could be helpful for multimodal analysis and purity assessment of CITE-seq datasets that target specific cell populations (e.g., B cells).

Human Hepatitis B Viral Infection Outcomes Are Linked to Naturally Occurring Variants of HLA-DOA That Have Altered Function.

HLA molecules of the MHC class II (MHCII) bind and present pathogen-derived peptides for CD4 T cell activation. Peptide loading of MHCII in the endosomes of cells is controlled by the interplay of the nonclassical MHCII molecules, HLA-DM (DM) and HLA-DO (DO). DM catalyzes peptide loading, whereas DO, an MHCII substrate mimic, prevents DM from interacting with MHCII, resulting in an altered MHCII-peptide repertoire and increased MHCII-CLIP. Although the two genes encoding DO (DOA and DOB) are considered nonpolymorphic, there are rare natural variants. Our previous work identified DOB variants that altered DO function. In this study, we show that natural variation in the DOA gene also impacts DO function. Using the 1000 Genomes Project database, we show that ∼98% of individuals express the canonical DOA*0101 allele, and the remaining individuals mostly express DOA*0102, which we found was a gain-of-function allele. Analysis of 25 natural occurring DOα variants, which included the common alleles, identified three null variants and one variant with reduced and nine with increased ability to modulate DM activity. Unexpectedly, several of the variants produced reduced DO protein levels yet efficiently inhibited DM activity. Finally, analysis of associated single-nucleotide polymorphisms genetically linked the DOA*0102 common allele, a gain-of-function variant, with human hepatitis B viral persistence. In contrast, we found that the DOα F114L null allele was linked with viral clearance. Collectively, these studies show that natural variation occurring in the human DOA gene impacts DO function and can be linked to specific outcomes of viral infections.

Polymorphic Immune Mechanisms Regulate Commensal Repertoire.

Environmental influences (infections and diet) strongly affect a host's microbiota. However, host genetics may influence commensal communities, as suggested by the greater similarity between the microbiomes of identical twins compared to non-identical twins. Variability of human genomes and microbiomes complicates the understanding of polymorphic mechanisms regulating the commensal communities. Whereas animal studies allow genetic modifications, they are sensitive to influences known as "cage" or "legacy" effects. Here, we analyze ex-germ-free mice of various genetic backgrounds, including immunodeficient and major histocompatibility complex (MHC) congenic strains, receiving identical input microbiota. The host's polymorphic mechanisms affect the gut microbiome, and both innate (anti-microbial peptides, complement, pentraxins, and enzymes affecting microbial survival) and adaptive (MHC-dependent and MHC-independent) pathways influence the microbiota. In our experiments, polymorphic mechanisms regulate only a limited number of microbial lineages (independently of their abundance). Our comparative analyses suggest that some microbes may benefit from the specific immune responses that they elicit.

Integrated genomic profiling expands clinical options for patients with cancer.

Genomic analysis of paired tumor-normal samples and clinical data can be used to match patients to cancer therapies or clinical trials. We analyzed 500 patient samples across diverse tumor types using the Tempus xT platform by DNA-seq, RNA-seq and immunological biomarkers. The use of a tumor and germline dataset led to substantial improvements in mutation identification and a reduction in false-positive rates. RNA-seq enhanced gene fusion detection and cancer type classifications. With DNA-seq alone, 29.6% of patients matched to precision therapies supported by high levels of evidence or by well-powered studies. This proportion increased to 43.4% with the addition of RNA-seq and immunotherapy biomarker results. Combining these data with clinical criteria, 76.8% of patients were matched to at least one relevant clinical trial on the basis of biomarkers measured by the xT assay. These results indicate that extensive molecular profiling combined with clinical data identifies personalized therapies and clinical trials for a large proportion of patients with cancer and that paired tumor-normal plus transcriptome sequencing outperforms tumor-only DNA panel testing.

Clinical validation of the tempus xT next-generation targeted oncology sequencing assay.

We developed and clinically validated a hybrid capture next generation sequencing assay to detect somatic alterations and microsatellite instability in solid tumors and hematologic malignancies. This targeted oncology assay utilizes tumor-normal matched samples for highly accurate somatic alteration calling and whole transcriptome RNA sequencing for unbiased identification of gene fusion events. The assay was validated with a combination of clinical specimens and cell lines, and recorded a sensitivity of 99.1% for single nucleotide variants, 98.1% for indels, 99.9% for gene rearrangements, 98.4% for copy number variations, and 99.9% for microsatellite instability detection. This assay presents a wide array of data for clinical management and clinical trial enrollment while conserving limited tissue.

Integrating RNA expression and visual features for immune infiltrate prediction.

Patient responses to cancer immunotherapy are shaped by their unique genomic landscape and tumor microenvironment. Clinical advances in immunotherapy are changing the treatment landscape by enhancing a patient's immune response to eliminate cancer cells. While this provides potentially beneficial treatment options for many patients, only a minority of these patients respond to immunotherapy. In this work, we examined RNA-seq data and digital pathology images from individual patient tumors to more accurately characterize the tumor-immune microenvironment. Several studies implicate an inflamed microenvironment and increased percentage of tumor infiltrating immune cells with better response to specific immunotherapies in certain cancer types. We developed NEXT (Neural-based models for integrating gene EXpression and visual Texture features) to more accurately model immune infiltration in solid tumors. To demonstrate the utility of the NEXT framework, we predicted immune infiltrates across four different cancer types and evaluated our predictions against expert pathology review. Our analyses demonstrate that integration of imaging features improves prediction of the immune infiltrate. Of note, this effect was preferentially observed for B cells and CD8 T cells. In sum, our work effectively integrates both RNA-seq and imaging data in a clinical setting and provides a more reliable and accurate prediction of the immune composition in individual patient tumors.

RNA Sequencing of the Tumor Microenvironment in Precision Cancer Immunotherapy.

RNA sequencing (RNA-seq) provides an efficient high-throughput technique to robustly characterize the tumor immune microenvironment (TME). The increasing use of RNA-seq in clinical and basic science settings provides a powerful opportunity to access novel therapeutic biomarkers in the TME. Advanced computational methods are making it possible to resolve the composition of the tumor immune infiltrate, infer the immunological phenotypes of those cells, and assess the immune receptor repertoire in RNA-seq data. These immunological characterizations have increasingly important implications for guiding immunotherapy use. Here, we highlight recent studies that demonstrate the potential utility of RNA-seq in clinical settings, review key computational methods used for characterizing the TME for precision cancer immunotherapy, and discuss important considerations in data interpretation and current technological limitations.