Search

Our study aims to investigate the dynamics of conventional memory T cells (Tconv) and regulatory T cells (Treg) following activation, and to explore potential differences between these two cell types. To achieve this, we developed advanced statistical mixed models and mathematical models of ordinary differential equations (ODE), which allowed us to transform post-vaccination immunological processes into mathematical formulas. These models were applied on in-house data from a de novo Hepatitis B vaccination trial. By accounting for inter- and intra-individual variability, our models provided good fits for both antigen-specific Tconv and Treg cells, overcoming the challenge of studying these complex processes. Our modeling approach provided a deeper understanding of the immunological processes underlying T cell development after vaccination. Specifically, our analysis revealed several important findings regarding the dynamics of Tconv and Treg cells, as well as their relationship to seropositivity for HSV-1 and EBV, and the dynamics of antibody response to vaccination. Firstly, our modeling indicated that Tconv dynamics suggest the existence of two T cell types, in contrast to Treg dynamics where only one T cell type is predicted. Secondly, we found that individuals who converted to a positive antibody response to the vaccine earlier had lower decay rates for both Tregs and Tconv cells, which may have important implications for the development of more effective vaccination strategies. Additionally, our modeling showed that HSV-1 seropositivity negatively influenced Tconv cell expansion after the second vaccination, while EBV seropositivity was associated with higher Treg expansion rates after vaccination. Overall, this study provides a critical foundation for understanding the dynamic processes underlying T cell development after vaccination.

Summary / AbstractEven high-performing machine learning models can have problems when deployed in a real-world setting if the data used to train and test the model contains biases. TCR–epitope binding prediction for novel epitopes is a very important but yet unsolved problem in immunology. In this article, we describe how the technique used to create negative data for the TCR–epitope interaction prediction task can lead to a strong bias and makes that the performance drops to random when tested in a more realistic scenario.

The varicella-zoster virus (VZV) infects over 95% of the population and establishes latency afterwards. Reactivation of VZV causes herpes zoster (HZ), commonly known as shingles, which presents as a painful rash in mostly the elderly and people with a weakened immune system. However, HZ might occur in otherwise healthy individuals too. In this study, we have studied the immune signature of HZ to better understand HZ’s pathophysiology. We provide a general overview of the antiviral state and the activation of innate and adaptive immune responses during HZ. Differential gene expression and gene ontology analyses revealed upregulation of several genes and host immune pathways during herpes zoster, especially related to type I IFN response but also related to adaptive immune responses. Intriguingly, no differences in gene expression were noted during convalescence between HZ patients and controls. Furthermore, we conducted the largest HLA association study on HZ to date using the UK Biobank and identified seven protective and four risk HLA alleles associated with the development of herpes zoster. These findings reveal key genes and pathways involved in the host immune response to symptomatic VZV reactivation and provide new molecular insights into the development of HZ.

The immune system acts as the intricate apparatus dedicated to mounting a defence that ensures host survival from microbial threats. To engage this faceted immune response and provide protection against infectious diseases, vaccinations are the critical tool developed. However, vaccine responses are governed by levels that when interrogated separately only explain a fraction of the immune reaction. To address this knowledge gap, we conducted a feasibility study to determine if multi-view modelling can aid in gaining actionable insights on response markers shared across populations, capture the immune system diversity, and disentangle confounders. We thus sought to assess this multi-view modelling capacity on the responsiveness to Hepatitis B virus (HBV) vaccination.Seroconversion to vaccine induced antibodies against HBV surface antigen (anti-HBs) in early-converters (n=21; 10IU/L). The multi-view data encompassed bulk RNA-seq, CD4+ T cell parameters (including T-cell receptor data), flow cytometry data, and clinical metadata (including age and gender). The modelling included testing single-view and multi-view joint dimensionality reductions. Multi-view joint dimensionality reduction out-performed single-view methods in terms of area under curve and balanced accuracy, confirming an increase in predictive power to be gained. The interpretation of the findings showed that age, gender, inflammation-related gene sets and pre-existing vaccine specific T-cells were associated with vaccination responsiveness.This multi-view dimensionality reduction approach complements the clinical seroconversion and all single modalities. Importantly, this modelling could identify what features predict HBV vaccine response. This methodology could be extended to other vaccination trials to identify key features regulating responsiveness.

Accurate molecular detection of resistance to bedaquiline, a core drug for the treatment of drug resistant tuberculosis, remains challenging. In this study, we investigated whether three sequence- and nine structure-based features describing the impact ofRv0678variants on the MmpR transcriptional repressor could predict the bedaquiline phenotype of isolates containingRv0678variants. Paired genotypic and phenotypic data was used to train a binary random forest classifier. The mean value of the individual features was similar for resistant and susceptible variants (p≥0.05). Leave one out cross validation showed predictability of bedaquiline resistance fromRv0678variants when using a binary classifier with a combination of sequence and structural features (ROC AUC = 0.766; f1 score = 0.746), but the performance was too low for clinically use. Evolutionary conservation of the affected residue was the most important individual feature (mean decrease in impurity = 0.226 ± 0.003) to discriminate bedaquiline resistance from susceptibility. Prediction of bedaquiline resistance was only possible when restricting the data to missense variants, as the selected features could not be applied toRv0678insertions and deletions. Additionally, prediction of bedaquiline resistance was only possible when restricting the data to isolates for which the phenotype was determined using Mycobacterial Growth Indicator Tubes, suggesting possible misclassification of the bedaquiline phenotype by other methods. The results of this study suggest that structural features describingRv0678missense variants could be used to predict bedaquiline resistance, albeit not yet at the performance level required for clinical practice.Author summaryGuidelines by the World Health Organization indicate that antibiotic resistant tuberculosis should be treated with bedaquiline, a newly discovered antibiotic. However, resistance to bedaquiline has already emerged and is spreading rapidly. More than 500 different variants have been discovered in the bedaquiline resistance genes in Mycobacterium tuberculosis. It is not known which DNA variants specifically drive bedaquiline resistance due to insufficiency of data. Here, we used a machine learning technique to predict whether a variant in theRv0678gene confers bedaquiline resistance. To this end, for each variant we defined features that describe the impact of the variant on the MmpR (encoded by theRv0678gene) protein structure. These features were then fed into a classification algorithm to predict bedaquiline resistance. We successfully developed a bedaquiline resistance prediction model, although performance was limited. We also found that accurate predictions were only possible when restricting the data to samples that were tested on the MGIT platform, a World Health Organization endorsed method to test for bedaquiline resistance. Our study shows that predicting bedaquiline resistance from protein structural features is feasible. Furthermore, our study provides new insights into phenotypic testing platforms to assess bedaquiline resistance.

BackgroundWhole-genome sequencing (WGS) is becoming an increasingly popular tool to study the population genetics and drug resistance ofPlasmodiumspp. However, the predominance of human DNA in a malaria patient blood sample requires time-consuming lab procedures to filter out human DNA or enrichPlasmodiumDNA. Here, we investigated the potential of adaptive sampling to enrich forPlasmodiumDNA while sequencing unenriched patient blood samples on a minION device.ResultsTo compare adaptive sampling versus regular sequencing, a dilution series consisting of 0% up to 100%P. falciparumDNA in human DNA was sequenced. Half of the flowcell channels were run in adaptive sampling mode, enriching for theP. falciparumreference genome, resulting in a 3.2 fold enrichment ofP. falciparumbases on average. Samples with a lower concentration of parasite DNA had a higher enrichment potential. We confirmed these findings by sequencing twoP. falciparumpatient blood samples with common levels of parasitaemia (0.1% and 0.2%). The estimated enrichment was 3.9 and 5.8, which was sufficient to cover at least 97% of theP. falciparumreference genome at a median depth of 20 (highest parasitaemia) or 5 (lowest parasitaemia). A comparison of 38 drug resistance variants (WHO) obtained via adaptive sequencing or Sanger sequencing showed a high concordance between the two methods, suggesting that the obtained sequencing data is of sufficient quality to address common clinical research questions for patients with parasitaemias of 0.1% and higher.ConclusionsOur results demonstrate that adaptive Nanopore sequencing has the potential to replace more time-consumingPlasmodium-enrichment protocols and sequence directly from patient blood, given further improvements in cost-efficiency.

A key analysis task in mass spectrometry proteomics is matching the acquired tandem mass spectra to their originating peptides by sequence database searching or spectral library searching. Machine learning is an increasingly popular post-processing approach to maximize the number of confident spectrum identifications that can be obtained at a given false discovery rate threshold. Here, we have integrated semi-supervised machine learning in the ANN-SoLo tool, an efficient spectral library search engine that is optimized for open modification searching to identify peptides with any type of post-translational modification. We show that machine learning rescoring boosts the number of spectra that can be identified for both standard searching and open searching, and we provide insights into relevant spectrum characteristics harnessed by the machine learning model. The semi-supervised machine learning functionality has now been fully integrated into ANN-SoLo, which is available as open source under the permissive Apache 2.0 license on GitHub athttps://github.com/bittremieux/ANN-SoLo.

The recognition of an epitope by a T-cell receptor (TCR) is crucial for eliminating pathogens and establishing immunological memory. Prediction of the binding of any TCR–epitope pair is still a challenging task, especially for novel epitopes, because the underlying patterns are largely unknown to domain experts and machine learning models. To achieve a deeper understanding of TCR–epitope interactions, we have used interpretable deep learning techniques to gain insights into the performance of TCR–epitope binding machine learning models. We demonstrate how interpretable AI techniques can be linked to the three-dimensional structure of molecules to offer novel insights into the factors that determine TCR affinity on a molecular level. Additionally, our results show the importance of using interpretability techniques to verify the predictions of machine learning models for challenging molecular biology problems where small hard-to-detect problems can accumulate to inaccurate results.

Ferroptosis is a lipid peroxidation-dependent mechanism of regulated cell death known to suppress tumor proliferation and progression. Although several genetic and protein hallmarks have been identified in ferroptotic cell death, it remains challenging to fully characterize ferroptosis signaling pathways and to find suitable biomarkers. Moreover, changes taking place in the epigenome of ferroptotic cells remain poorly studied. In this context, we aimed to investigate the role of chromatin remodeler forkhead box protein A1 (FOXA1) in RSL3-treated multiple myeloma cells because, similar to ferroptosis, this transcription factor has been associated with changes in the lipid metabolism, DNA damage, and epithelial-to-mesenchymal transition (EMT). RNA sequencing and Western blot analysis revealed that FOXA1 expression is consistently upregulated upon ferroptosis induction in different in vitro and in vivo disease models. In silico motif analysis and transcription factor enrichment analysis further suggested that ferroptosis-mediated FOXA1 expression is orchestrated by specificity protein 1 (Sp1), a transcription factor known to be influenced by lipid peroxidation. Remarkably, FOXA1 upregulation in ferroptotic myeloma cells did not alter hormone signaling or EMT, two key downstream signaling pathways of FOXA1. CUT&RUN genome-wide transcriptional binding site profiling showed that GPX4-inhibition by RSL3 triggered loss of binding of FOXA1 to pericentromeric regions in multiple myeloma cells, suggesting that this transcription factor is possibly involved in genomic instability, DNA damage, or cellular senescence under ferroptotic conditions.Abstract Figure

Aneuploidy causes broad-scale disruption in the stochiometric balances of transcripts, proteins, and metabolites. These disruptions often lead to detrimental effects for the organism, but for some organisms and in certain environments, it can also cause a fitness advantage. Understanding the complex trade-off between aneuploidy fitness gains and losses requires a systems biological comprehension of its molecular impact. The protozoan parasite Leishmania provides a unique study angle to this problem, as the pathogen lacks transcriptional regulation of individual protein-coding genes and has an unusually high tolerance for aneuploidy. Here, we present the first integrated analysis of the genome, transcriptome, proteome, and metabolome of highly aneuploid Leishmania donovani strains. This 4 layer multi-omics analysis unambiguously establishes that despite this tolerance, aneuploidy in Leishmania globally and drastically impacts the transcriptome and proteome of affected chromosomes, ultimately explaining the degree of metabolic differences between strains. This impact appears to be present throughout the parasite’s life cycle as we observed it in both its proliferative and infectious life stages. We show that, through dosage compensation, protein complex subunits, secreted proteins, and non-cytoplasmic proteins responded less or even not at all to aneuploidy. In contrast to other Eukaryotes, we did not observe the widespread regulation at the transcript level that typically modulate some of the negative effects of aneuploidy. Instead, Leishmania features a surprisingly high number of regulated proteins encoded by non-aneuploid chromosomes, suggesting that aneuploidy results in extensive post-transcriptionally modulation of protein levels. Our results provide the foundation for the integrated molecular understanding of aneuploidy in Leishmania. They also highlight that the parasite is an attractive model to study processes of post-transcriptional modulation of protein levels in the context of aneuploidy and beyond.

The T-cell receptor (TCR) determines the specificity of a T-cell towards an epitope. As of yet, the rules for antigen recognition remain largely undetermined. Current methods for grouping TCRs according to their epitope specificity remain limited in performance and scalability. Multiple methodologies have been developed, but all of them fail to efficiently cluster large data sets exceeding 1 million sequences. To account for this limitation, we developed clusTCR, a rapid TCR clustering alternative that efficiently scales up to millions of CDR3 amino acid sequences. Benchmarking comparisons revealed similar accuracy of clusTCR with other TCR clustering methods. clusTCR offers a drastic improvement in clustering speed, which allows clustering of millions of TCR sequences in just a few minutes through efficient similarity searching and sequence hashing.clusTCR was written in Python 3. It is available as an anaconda package (https://anaconda.org/svalkiers/clustcr) and on github (https://github.com/svalkiers/clusTCR).

Susceptibility for leishmaniasis is largely dependent on genetic- and immune factors of the host. Despite the previously described association of human leukocyte antigen (HLA) gene cluster variants as genetic susceptibility factors, little is known on the mechanisms that mediate these associations. To characterize the functionality underpinning these associations between HLA and disease, we predicted the epitope binding repertoires for all known leishmaniasis-associated HLA variants collected in a thorough literature review. We identified several amino acid polymorphisms in the HLA sequences that distinguished protective-from risk-associated HLA-DRB1 alleles. Proteome-wide and multi-species T cell epitope binding predictions were carried out across these alleles, enabling us to map the effects on the epitope binding repertoires. The protective-associated HLA-DRB1 alleles were characterized by common binding core motifs, which map to the identified amino acid polymorphisms. These results strongly suggest that polymorphism in the HLA region, resulting in differential antigen presentation, affects the association between HLA and leishmaniasis disease development. Finally, we established a valuable open-access resource of putative epitopes, of which a set of 14 HLA-unrestricted strong-binding epitopes, conserved across species, were prioritized for further epitope discovery in the search for novel subunit-based vaccines.

RationaleAdvanced algorithmic solutions are necessary to process the ever increasing amounts of mass spectrometry data that is being generated. Here we describe the falcon spectrum clustering tool for efficient clustering of millions of MS/MS spectra.Methodsfalcon succeeds in efficiently clustering large amounts of mass spectral data using advanced techniques for fast spectrum similarity searching. First, high-resolution spectra are binned and converted to low-dimensional vectors using feature hashing. Next, the spectrum vectors are used to construct nearest neighbor indexes for fast similarity searching. The nearest neighbor indexes are used to efficiently compute a sparse pairwise distance matrix without having to exhaustively perform all pairwise spectrum comparisons within the relevant precursor mass tolerance. Finally, density-based clustering is performed to group similar spectra into clusters.ResultsSeveral state-of-the-art spectrum clustering tools were evaluated using a large draft human proteome dataset consisting of 25 million spectra, indicating that alternative tools produce clustering results with different characteristics. Notably, falcon generates larger highly pure clusters than alternative tools, leading to a larger reduction in data volume without the loss of relevant information for more efficient downstream processing.Conclusionsfalcon is a highly efficient spectrum clustering tool. It is publicly available as open source under the permissive BSD license at https://github.com/bittremieux/falcon.

Despite the general agreement on the importance of T cells during SARS-CoV-2 infection, the clinical impact of specific and cross-reactive T-cell responses remains uncertain, while this knowledge may indicate how to adjust vaccines and maintain robust long-term protection against continuously emerging variants. To characterize CD8+ T-cell response to epitopes unique to SARS-CoV-2 (SC-unique) or shared with other coronaviruses (CoV-common), we trained a large number of TCR-epitope recognition models for MHC-I-presented SARS-CoV-2 epitopes from publicly available data. Applying those models to longitudinal COVID-19 TCR repertoires of critical and non-critical COVID-19 patients, we discovered that notwithstanding comparable CD8+ T-cell depletion and the sizes of putative CoV-common CD8+ TCR repertoires in all symptomatic patients at the initial stage of the disease, the temporal dynamics of putative SC2-unique TCRs differed depending on the disease severity. Only non-critical patients had developed large and diverse SC2-unique CD8+ T-cell response by the second week of the disease. Additionally, only this patient group demonstrated redundancy in CD8+ TCRs putatively recognizing unique and common SARS-CoV-2 epitopes. Our findings thus emphasize the role of thede novoCD8+ T-cell response and support the argument against the clinical benefit of pre-existing cross-reactive CD8+ T cells. Now, the analytical framework of this study can not only be employed to track specific and cross-reactive SARS-CoV-2 CD8+ T cells in any TCR repertoire but also be generalized to more epitopes and be employed for adaptive immune response assessment and monitoring to inform public health decisions.

The prediction of epitope recognition by T-cell receptors (TCRs) has seen many advancements in recent years, with several methods now available that can predict recognition for a specific set of epitopes. However, the generic case of evaluating all possible TCR-epitope pairs remains challenging, mainly due to the high diversity of the interacting sequences and the limited amount of currently available training data. In this work, we provide an overview of the current state of this unsolved problem. First, we examine appropriate validation strategies to accurately assess the generalization performance of generic TCR-epitope recognition models when applied to both known and novel epitopes. In addition, we present a novel feature representation approach which we call ImRex (interaction map recognition). This approach is based on the pairwise combination of physicochemical properties of the individual amino acids in the CDR3 and epitope sequences, which provides a convolutional neural network with the combined representation of both sequences. Lastly, we highlight various challenges that are particular to TCR-epitope data and that can adversely affect model performance. These include the issue of selecting negative data, the imbalanced epitope distribution of curated TCR-epitope datasets, and the potential exchangeability of TCR alpha and beta chains. Our results indicate that while extrapolation to novel epitopes remains a difficult challenge, ImRex makes this feasible for a subset of epitopes that are not too dissimilar from the training data. We show that appropriate feature engineering methods and rigorous benchmark standards are required to create and validate TCR-epitope predictive models.

Open modification searching (OMS) is a powerful search strategy to identify peptides with any type of modification. OMS works by using a very wide precursor mass window to allow modified spectra to match against their unmodified variants, after which the modification types can be inferred from the corresponding precursor mass differences. A disadvantage of this strategy, however, is the large computational cost, because each query spectrum has to be compared against a multitude of candidate peptides. We have previously introduced the ANN-SoLo tool for fast and accurate open spectral library searching. ANN-SoLo uses approximate nearest neighbor indexing to speed up OMS by selecting only a limited number of the most relevant library spectra to compare to an unknown query spectrum. Here we demonstrate how this candidate selection procedure can be further optimized using graphics processing units. Additionally, we introduce a feature hashing scheme to convert high-resolution spectra to low-dimensional vectors. On the basis of these algorithmic advances, along with low-level code optimizations, the new version of ANN-SoLo is up to an order of magnitude faster than its initial version. This makes it possible to efficiently perform open searches on a large scale to gain a deeper understanding about the protein modification landscape. We demonstrate the computational efficiency and identification performance of ANN-SoLo based on a large data set of the draft human proteome. ANN-SoLo is implemented in Python and C++. It is freely available under the Apache 2.0 license at https://github.com/bittremieux/ANN-SoLo. This document is the unedited Author's version of a Submitted Work that was subsequently accepted for publication in the Journal of Proteome Research, copyright © American Chemical Society after peer review. To access the final edited and published work see https://pubs.acs.org/articlesonrequest/AOR-TCA25r6IZJVZDTDN337H.

Data analysis for metabolomics is undergoing rapid progress thanks to the proliferation of novel tools and the standardization of existing workflows. However, as datasets and experiments continue to increase in size and complexity, standardized workflows are often not sufficient. In addition, as the ground truth for metabolomics experiments is intrinsically unknown, there is no way to critically evaluate the performance of tools. Here, we investigate the problem of dynamic multi-class metabolomics experiments using a simulated dataset with a known ground truth and evaluate the performance of tinderesting, a new and intuitive tool based on gathering expert knowledge to be used in machine learning, and compare it to EDGE, a statistical method for sequence data. This paper presents three novel outcomes. First we present a way to simulate dynamic metabolomics data with a known ground truth based on ordinary differential equations. This method is made available through the MetaboLouise R package. Second, we show that the EDGE tool, originally developed for genomics data analysis, is highly performant in analyzing dynamic case vs control metabolomics data. Last, we introduce the tinderesting method to analyse more complex dynamic metabolomics experiments that performs on par with statistical methods. This tool consists of a Shiny app for collecting expert knowledge, which in turn is used to train a machine learning model to emulate the decision process of the expert. This approach does not replace traditional data analysis workflows for metabolomics, but can provide additional information, improved performance or easier interpretation of results. The advantage is that the tool is agnostic to the complexity of the experiment, and thus is easier to use in advanced setups. All code for the presented analysis, MetaboLouise and tinderesting are freely available.

High-throughput T cell receptor (TCR) sequencing allows the characterization of an individual’s TCR repertoire and directly query their immune state. However, it remains a non-trivial task to couple these sequenced TCRs to their antigenic targets. In this paper, we present a novel strategy to annotate full TCR sequence repertoires. The strategy is based on a machine learning algorithm to learn the TCR patterns common to the recognition of a specific epitope. These results are then combined with a statistical analysis to evaluate the occurrence of specific epitope-reactive TCR sequences per epitope in repertoire data. In this manner, we can directly study the capacity of full TCR repertoires to target specific epitopes of the relevant vaccines or pathogens. We demonstrate the usability of this approach on three independent datasets related to vaccine monitoring and infectious disease diagnostics by independently identifying the epitopes that are targeted by the TCR repertoire. The developed method is freely available as a web tool for academic use at tcrex.biodatamining.be.

The T-cell receptor is responsible for recognizing potentially harmful epitopes presented on cell surfaces. The binding rules that govern this recognition between receptor and epitope is currently an unsolved problem, yet one of great interest. Several methods have been proposed recently to perform supervised classification of T-cell receptor sequences, but this requires known examples of T-cell sequences for a given epitope. Here we study the viability of various methods to perform unsupervised clustering of distinct T-cell receptor sequences and how these clusters relate to their target epitope. The goal is to provide an overview of the performance of various distance metrics on two large independent T-cell receptor sequence data sets. Our results confirm the presence of structural distinct T-cell groups that target identical epitopes. In addition, we put forward several recommendations to perform T-cell receptor sequence clustering.

Current T-cell epitope prediction tools are a valuable resource in designing targeted immunogenicity experiments. They typically focus on, and are able to, accurately predict peptide binding and presentation by major histocompatibility complex (MHC) molecules on the surface of antigen-presenting cells. However, recognition of the peptide-MHC complex by a T-cell receptor is often not included in these tools. We developed a classification approach based on random forest classifiers to predict recognition of a peptide by a T-cell and discover patterns that contribute to recognition. We considered two approaches to solve this problem: (1) distinguishing between two sets of T-cell receptors that each bind to a known peptide and (2) retrieving T-cell receptors that bind to a given peptide from a large pool of T-cell receptors. Evaluation of the models on two HIV-1, B*08-restricted epitopes reveals good performance and hints towards structural CDR3 features that can determine peptide immunogenicity. These results are of particularly importance as they show that prediction of T-cell epitope and T-cell epitope recognition based on sequence data is a feasible approach. In addition, the validity of our models not only serves as a proof of concept for the prediction of immunogenic T-cell epitopes but also paves the way for more general and high performing models.