scholarly journals DeepBCR: Deep learning framework for cancer-type classification and binding affinity estimation using B cell receptor repertoires

2019 ◽  
Author(s):  
Xihao Hu ◽  
X Shirley Liu
Keyword(s):  
Deep Learning ◽  
B Cell ◽  
Binding Affinity ◽  
Cell Receptor ◽  
B Cell Receptors ◽  
Cancer Type ◽  
Binding Affinities ◽  
Learning Framework ◽  
Kd Values ◽  
Type Classification

AbstractWe developed a deep learning framework to model the binding specificity of B-cell receptors (BCRs). The DeepBCR framework can predict the cancer type from a repertoire of BCRs and estimate the binding affinity of a single BCR. We designed a peptide encoding network that includes an amino acid encoding layer, k-mer motif layer, and immunoglobulin isotype layer, and used transfer learning to reduce parameters and over-fitting. When we applied the framework to evaluate the three commercial anti-PD1 drugs (Opdivo, Keytruda, and Libtayo), the predicted binding affinities correlate with the real affinities measured in kD values. This validates the prediction and indicates that we can use the framework to select strong antigen-specific binders.

scholarly journals Convergent selection in antibody repertoires is revealed by deep learning

2020 ◽  
Author(s):  
Simon Friedensohn ◽  
Daniel Neumeier ◽  
Tarik A Khan ◽  
Lucia Csepregi ◽  
Cristina Parola ◽  
...  
Keyword(s):  
Deep Learning ◽  
T Cell ◽  
B Cell ◽  
Recombinant Expression ◽  
Cell Receptor ◽  
Support Vector ◽  
Cell Receptors ◽  
Sequence Patterns ◽  
Antibody Repertoires ◽  
Selection Of

SUMMARYAdaptive immunity is driven by the ability of lymphocytes to undergo V(D)J recombination and generate a highly diverse set of immune receptors (B cell receptors/secreted antibodies and T cell receptors) and their subsequent clonal selection and expansion upon molecular recognition of foreign antigens. These principles lead to remarkable, unique and dynamic immune receptor repertoires1. Deep sequencing provides increasing evidence for the presence of commonly shared (convergent) receptors across individual organisms within one species2-4. Convergent selection of specific receptors towards various antigens offers one explanation for these findings. For example, single cases of convergence have been reported in antibody repertoires of viral infection or allergy5-8. Recent studies demonstrate that convergent selection of sequence motifs within T cell receptor (TCR) repertoires can be identified on an even wider scale9,10. Here we report that there is extensive convergent selection in antibody repertoires of mice for a range of protein antigens and immunization conditions. We employed a deep learning approach utilizing variational autoencoders (VAEs) to model the underlying process of B cell receptor (BCR) recombination and assume that the data generation follows a Gaussian mixture model (GMM) in latent space. This provides both a latent embedding and cluster labels that group similar sequences, thus enabling the discovery of a multitude of convergent, antigen-associated sequence patterns. Using a linear, one-versus-all support vector machine (SVM), we confirm that the identified sequence patterns are predictive of antigenic exposure and outperform predictions based on the occurrence of public clones. Recombinant expression of both natural and in silico-generated antibodies possessing convergent patterns confirms their binding specificity to target antigens. Our work highlights to which extent convergence in antibody repertoires can occur and shows how deep learning can be applied for immunodiagnostics and antibody discovery and engineering.

scholarly journals Autoreactive B Cell Receptors Mimic Autonomous Pre-B Cell Receptor Signaling and Induce Proliferation of Early B Cells

Immunity ◽  
2008 ◽  
Vol 29 (6) ◽  
pp. 912-921 ◽  
Author(s):  
Fabian Köhler ◽  
Eva Hug ◽  
Cathrin Eschbach ◽  
Sonja Meixlsperger ◽  
Elias Hobeika ◽  
...  
Keyword(s):  
B Cells ◽  
B Cell ◽  
Cell Receptor ◽  
B Cell Receptor ◽  
B Cell Receptors ◽  
Receptor Signaling ◽  
Cell Receptors ◽  
B Cell Receptor Signaling ◽  
Cell Receptor Signaling

scholarly journals Structurally distinct endocytic pathways for B cell receptors in B lymphocytes

2020 ◽  
Vol 31 (25) ◽  
pp. 2826-2840
Author(s):  
Aleah D. Roberts ◽  
Thaddeus M. Davenport ◽  
Andrea M. Dickey ◽  
Regina Ahn ◽  
Kem A. Sochacki ◽  
...  
Keyword(s):  
B Cell ◽  
Cell Receptor ◽  
B Cell Receptor ◽  
B Cell Receptors ◽  
Cortical Actin ◽  
Receptor Endocytosis ◽  
Low Concentrations ◽  
High Concentrations ◽  
Receptor Clusters ◽  
Endocytic Pathways

We report structural and mechanistic differences in B cell receptor endocytosis at high and low concentrations of antigen. We propose that the mechanism of endocytosis switches to accommodate large receptor clusters formed when cells encounter high concentrations of antigen. This mechanism is regulated by the dynamics of the cortical actin cytoskeleton.

Intratumoral T cell clonality and survival in a randomized phase II study of preoperative autophagy inhibition in combination with gemcitabine and nab-paclitaxel treatment in patients with resectable pancreatic cancer.

2021 ◽  
Vol 39 (15_suppl) ◽  
pp. e16001-e16001
Author(s):  
Pranav Murthy ◽  
Daniel Weber ◽  
Sagar N Sharma ◽  
Aatur D. Singhi ◽  
Nathan Bahary ◽  
...  
Keyword(s):  
Clinical Trial ◽  
T Cells ◽  
T Cell ◽  
B Cell ◽  
T Cell Receptor ◽  
Cd8 T Cells ◽  
Cell Receptor ◽  
B Cell Receptors ◽  
Randomized Phase Ii ◽  
Autophagy Inhibition

e16001 Background: Autophagy is a cell survival mechanism that is upregulated in pancreatic ductal adenocarcinoma (PDAC). PDAC autophagy results in an altered metabolic phenotype that promotes tumor progression, chemotherapeutic resistance, and immune evasion. Methods: We previously completed a randomized phase II clinical trial of preoperative gemcitabine-nab-paclitaxel with (PGH n = 34) and without (PG, n = 30) autophagy inhibition in patients with resectable and borderline resectable PDAC, which demonstrated increased Evans Grade histopathologic and serum CA 19-9 response with autophagy inhibition (IRB 13-074, NCT01128296 ). Utilizing the resected FFPE tumor specimens from evaluable patients, we completed paired multiplex immunohistochemistry (CD4, CD8, FOXP3, CD20, CD68, Pan-CK) and T & B cell receptor RNA sequencing to assess the intratumoral adaptive immune response and correlates of outcome. Results: Autophagy inhibition increased the number of infiltrating CD8 T cells (1133±490 vs 712±460 average cells per high power field, p = 0.01), CD8:CD20 ratio (2.22±3.1 vs 0.96±1.1, p = 0.02) and reduced the CD4:CD8 ratio (2.04±0.87 vs 3.01±2.09, p = 0.03). No effect was observed on the number of immature or mature germinal center-like tertiary lymphoid structures (TLS), though the number of TLS correlated with increased infiltration of CD4 T cells (r = 0.40, p < 0.001), T-regulatory cells (r = 0.26, p = 0.03) and CD20 B cells (r = 0.65, p < 0.001). Although the total number of productive T and B cell receptors increased with autophagy inhibition (167217±105961 vs 97339±5,1628, p = 0.02), no apparent effects were observed on Vαβ TCR or BCR IgH, Igκ, Igλ clonality. Independent of treatment, intratumoral CD8 counts were associated with an improved CA 19-9 response (r = 0.32, p = 0.04) and in a subset of short term ( < 2 years, n = 17) and long term ( > 4 years, n = 10) survivors (LTS), a lowered CD4:CD8 ratio was identified in LTS (1.83±0.63 vs 2.8±0.90, p = 0.01). Dominance of B cell receptors was a prominent feature of the immune repertoire in all patients (average expression: Vα 0.6%, Vβ 0.8%, IgH 18.9%, Igκ 32.3%, Igλ 47.2%) with an IgA skewed immunoglobulin class switching (mean 63% of all BCRs). Increased αβ T cell receptor clonality above the median level was associated with a CA 19-9 response (r = 0.37, p = 0.06) and greater overall survival (median OS 38.3 vs 19.3 months, p = 0.02), indicative of possible tumor specific clonal expansion. Conclusions: Preoperative autophagy inhibition increased the number of tumor infiltrating CD8 T cells in patients with localized pancreatic cancer. Intratumoral αβ T cell receptor clonality was associated with CA 19-9 response and improved overall survival. Combination treatment regimens increasing PDAC specific CD8 responses are warranted. Clinical trial information: NCT01978184.

scholarly journals NOVEL AGENTS AND EMERGING STRATEGIES FOR TARGETING THE B-CELL RECEPTOR PATHWAY IN CLL

2012 ◽  
Vol 4 (1) ◽  
pp. e2012067 ◽  
Author(s):  
Dimitar Efremov ◽  
Adrian Wiestner ◽  
Luca Laurenti
Keyword(s):  
B Cell ◽  
Leukemic Cell ◽  
Cell Receptor ◽  
Therapeutic Agents ◽  
B Cell Receptor ◽  
Lymphocytic Leukemia ◽  
B Cell Receptors ◽  
Clinical Responses ◽  
Preclinical And Clinical Studies ◽  
Marked Dependence

Chronic lymphocytic leukemia (CLL) is a disease of malignant CD5+ B lymphocytes that are characterized by frequent expression of autoreactive B-cell receptors (BCRs) and marked dependence on microenvironmental signals for proliferation and survival. Among the latter, signals propagated through the BCR are believed to play a key role in leukemia initiation, maintenance and evolution. Drugs that can disrupt these signals have recently emerged as potential therapeutic agents in CLL and several of them are currently being evaluated in clinical trials. Particularly promising clinical responses have been obtained with inhibitors of the kinases SYK, BTK, and PI3Kδ, which function by blocking BCR signal transduction. In addition, recent studies focusing on the phosphatase PTPN22, which is involved in the pathogenesis of multiple autoimmune diseases and is markedly overexpressed in CLL cells, suggest that it may be possible in the future to develop strategies that will selectively reprogram BCR survival signals into signals that induce leukemic cell death. This review focuses on the biological basis behind these strategies and highlights some of the most promising BCR-targeting agents in ongoing preclinical and clinical studies.  

scholarly journals B cell receptor ligation induces display of V-region peptides on MHC class II molecules to T cells

2019 ◽  
Vol 116 (51) ◽  
pp. 25850-25859 ◽  
Author(s):  
Peter Csaba Huszthy ◽  
Ramakrishna Prabhu Gopalakrishnan ◽  
Johanne Tracey Jacobsen ◽  
Ole Audun Werner Haabeth ◽  
Geir Åge Løset ◽  
...  
Keyword(s):  
T Cells ◽  
B Cells ◽  
B Cell ◽  
Mhc Class Ii ◽  
Cell Receptor ◽  
B Cell Receptor ◽  
B Cell Receptors ◽  
V Region ◽  
Receptor Ligation

The B cell receptors (BCRs) for antigen express variable (V) regions that are enormously diverse, thus serving as markers on individual B cells. V region-derived idiotypic (Id) peptides can be displayed as pId:MHCII complexes on B cells for recognition by CD4+T cells. It is not known if naive B cells spontaneously display pId:MHCII in vivo or if BCR ligation is required for expression, thereby enabling collaboration between Id+B cells and Id-specific T cells. Here, using a mouse model, we show that naive B cells do not express readily detectable levels of pId:MHCII. However, BCR ligation by Ag dramatically increases physical display of pId:MHCII, leading to activation of Id-specific CD4+T cells, extrafollicular T–B cell collaboration and some germinal center formation, and production of Id+IgG. Besides having implications for immune regulation, the results may explain how persistent activation of self-reactive B cells induces the development of autoimmune diseases and B cell lymphomas.

scholarly journals DeepTCR is a deep learning framework for revealing sequence concepts within T-cell repertoires

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
John-William Sidhom ◽  
H. Benjamin Larman ◽  
Drew M. Pardoll ◽  
Alexander S. Baras
Keyword(s):  
Deep Learning ◽  
T Cell ◽  
Adaptive Immune System ◽  
Cell Receptor ◽  
Rna Seq ◽  
Sequencing Data ◽  
Learning Framework ◽  
Adaptive Immune ◽  
Gene Usage

AbstractDeep learning algorithms have been utilized to achieve enhanced performance in pattern-recognition tasks. The ability to learn complex patterns in data has tremendous implications in immunogenomics. T-cell receptor (TCR) sequencing assesses the diversity of the adaptive immune system and allows for modeling its sequence determinants of antigenicity. We present DeepTCR, a suite of unsupervised and supervised deep learning methods able to model highly complex TCR sequencing data by learning a joint representation of a TCR by its CDR3 sequences and V/D/J gene usage. We demonstrate the utility of deep learning to provide an improved ‘featurization’ of the TCR across multiple human and murine datasets, including improved classification of antigen-specific TCRs and extraction of antigen-specific TCRs from noisy single-cell RNA-Seq and T-cell culture-based assays. Our results highlight the flexibility and capacity for deep neural networks to extract meaningful information from complex immunogenomic data for both descriptive and predictive purposes.

scholarly journals Wide and Deep Learning for Automatic Cell Type Identification

2020 ◽  
Author(s):  
Christopher M. Wilson ◽  
Brooke L. Fridley ◽  
José Conejo-Garcia ◽  
Xuefeng Wang ◽  
Xiaoqing Yu
Keyword(s):  
Deep Learning ◽  
Prediction Accuracy ◽  
Predictive Performance ◽  
Cancer Type ◽  
Cell Type ◽  
Final Model ◽  
Biologically Relevant ◽  
A Cell ◽  
Classification Prediction ◽  
Type Classification

AbstractCell type classification is an important problem in cancer research, especially with the advent of single cell technologies. Correctly identifying cells within the tumor microenvironment can provide oncologists with a snapshot of how a patient’s immune system is reacting to the tumor. Wide deep learning (WDL) is an approach to construct a cell-classification prediction model that can learn patterns within high-dimensional data (deep) and ensure that biologically relevant features (wide) remain in the final model. In this paper, we demonstrate that the use of regularization can prevent overfitting and adding a wide component to a neural network can result in a model with better predictive performance. In particular, we observed that a combination of dropout and ℓ2 regularization can lead to a validation loss function that does not depend on the number of training iterations and does not experience a significant decrease in prediction accuracy compared to models with ℓ1, dropout, or no regularization. Additionally, we show WDL can have superior classification accuracy when the training and testing of a model is completed data on that arise from the same cancer type, but from different platforms. More specifically, WDL compared to traditional deep learning models can substantially increase the overall cell type prediction accuracy (41 to 90%) and T-cell sub-types (CD4: 0 to 76%, and CD8: 61 to 96%) when the models were trained using melanoma data obtained from the 10X platform and tested on basal cell carcinoma data obtained using SMART-seq.

Sign in / Sign up

Export Citation Format

Share Document