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

Abstract

Deep 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.

Unconstrained generation of synthetic antibody-antigen structures to guide machine learning methodology for real-world antibody specificity prediction

“…Then, an accuracy of 100% will be impossible, and a formulation with predicting one of the possible epitopes as loss function, or using a Multiple Instance Learning formalism could help to learn from datasets with higher levels of cross-reactivity (93,94).…”

“…However, performing paratope-epitope prediction with datasets containing more antigens (mutated for instance) might require a more complex ML task formulation, since more answers will be possible for the same paratope (multiple epitopes for the same paratope, when a CDRH3 was binding different antigens with a different conformational epitope). Then, an accuracy of 100% will be impossible, and a formulation with predicting one of the possible epitopes as loss function, or using a Multiple Instance Learning formalism could help to learn from datasets with higher levels of cross-reactivity (93,94).…”

Deep learning identifies antigenic determinants of severe SARS-CoV-2 infection within T-cell repertoires

“…Deep learning models were implemented with DeepTCR (v2.0.10) python package (https:// pypi. org/ proje ct/ DeepT CR/) 26 .…”

Biological controls for standardization and interpretation of adaptive immune receptor repertoire profiling

“…Furthermore, whether individuals used controls or did not, they appeared to agree on the types of issues in analyzing and interpreting AIRR-seq data that would benefit from the use of controls - most importantly measuring and controlling for sample quality, assay sensitivity and specificity, and calibration for the quantification of clonal size. Also, with the progress regarding antigen-specificity inference using computational tools ( Glanville et al, 2017 ; Huang et al, 2020 ; Jokinen et al, 2021 ; Akbar et al, 2021 ; Hayashi et al, 2021 ; Richardson et al, 2021 ; Galson et al, 2015 ; Shomuradova et al, 2020 ; Chronister et al, 2021 ; Sidhom et al, 2021 ; Pogorelyy et al, 2019 ; Dash et al, 2017 ) or more conventionally through technologically challenging using antigen-binding approaches, including single-cell ( Johnson et al, 2020 ; Fuchs et al, 2019 ), having controls would be of major interest to ensure the accuracy of the TR or IG identified. It is unlikely that a single control can fulfill all of these needs across all methods and applications.…”