Vaxi-DL: A web-based deep learning server to identify potential vaccine candidates

Leveraging artificial intelligence in vaccine development: A narrative review

Vaccine development stands as a cornerstone of public health efforts, pivotal in curbing infectious diseases and reducing global morbidity and mortality. However, traditional vaccine development methods are often time-consuming, costly, and inefficient. The advent of artificial intelligence (AI) has ushered in a new era in vaccine design, offering unprecedented opportunities to expedite the process. This narrative review explores the role of AI in vaccine development, focusing on antigen selection, epitope prediction, adjuvant identification, and optimization strategies. AI algorithms, including machine learning and deep learning, leverage genomic data, protein structures, and immune system interactions to predict antigenic epitopes, assess immunogenicity, and prioritize antigens for experimentation. Furthermore, AI-driven approaches facilitate the rational design of immunogens and the identification of novel adjuvant candidates with optimal safety and efficacy profiles. Challenges such as data heterogeneity, model interpretability, and regulatory considerations must be addressed to realize the full potential of AI in vaccine development. Integrating emerging technologies, such as single-cell omics and synthetic biology, promises to enhance vaccine design precision and scalability. This review underscores the transformative impact of AI on vaccine development and highlights the need for interdisciplinary collaborations and regulatory harmonization to accelerate the delivery of safe and effective vaccines against infectious diseases.

Subtractive Proteomics and Reverse-Vaccinology Approaches for Novel Drug Target Identification and Chimeric Vaccine Development against Bartonella henselae Strain Houston-1

Bartonella henselae is a Gram-negative bacterium causing a variety of clinical symptoms, ranging from cat-scratch disease to severe systemic infections, and it is primarily transmitted by infected fleas. Its status as an emerging zoonotic pathogen and its capacity to persist within host erythrocytes and endothelial cells emphasize its clinical significance. Despite progress in understanding its pathogenesis, limited knowledge exists about the virulence factors and regulatory mechanisms specific to the B. henselae strain Houston-1. Exploring these aspects is crucial for targeted therapeutic strategies against this versatile pathogen. Using reverse-vaccinology-based subtractive proteomics, this research aimed to identify the most antigenic proteins for formulating a multi-epitope vaccine against the B. henselae strain Houston-1. One crucial virulent and antigenic protein, the PAS domain-containing sensor histidine kinase protein, was identified. Subsequently, the identification of B-cell and T-cell epitopes for the specified protein was carried out and the evaluated epitopes were checked for their antigenicity, allergenicity, solubility, MHC binding capability, and toxicity. The filtered epitopes were merged using linkers and an adjuvant to create a multi-epitope vaccine construct. The structure was then refined, with 92.3% of amino acids falling within the allowed regions. Docking of the human receptor (TLR4) with the vaccine construct was performed and demonstrated a binding energy of -1047.2 Kcal/mol with more interactions. Molecular dynamic simulations confirmed the stability of this docked complex, emphasizing the conformation and interactions between the molecules. Further experimental validation is necessary to evaluate its effectiveness against B. henselae.

AI-based pipeline for early screening of lung cancer: integrating radiology, clinical, and genomics data

Background

The prognosis of lung carcinoma has changed since the discovery of molecular targets and their specific drugs. Somatic Epidermal Growth Factor Receptor (EGFR) mutations have been reported in lung carcinoma, and these mutant proteins act as substrates for targeted therapies. However, in a resource-constrained country like India, panel-based next-generation sequencing cannot be made available to the population at large. Additional challenges such as adequacy of tissue in case of lung core biopsies and locating suitable tumour tissues as a result of innate intratumoral heterogeneity indicate the necessity of an AI-based end-to-end pipeline capable of automatically detecting and learning more effective lung nodule features from CT images and predicting the probability of the EGFR-mutant. This will help the oncologists and patients in resource-limited settings to achieve near-optimal care and appropriate therapy.

Methods

The EGFR gene sequencing and CT imaging data of 2277 patients with lung carcinoma were included from three cohorts in India and a White population cohort collected from TCIA. Another cohort LIDC-IDRI was used to train the AIPS-Nodule (AIPS-N) model for automatic detection and characterisation of lung nodules. We explored the value of combining the results of the AIPS-N with the clinical factors in the AIPS-Mutation (AIPS-M) model for predicting EGFR genotype, and it was evaluated by area under the curve (AUC).

Findings

AIPS-N achieved an average AP50 of 70.19% in detecting the location of nodules within the lung region of interest during validation and predicted the score of five lung nodule properties. The AIPS-M machine learning (ML) and deep learning (DL) models achieved AUCs ranging from 0.587 to 0.910.

Interpretation

The AIPS suggests that CT imaging combined with a fully automated lung-nodule analysis AI system can predict EGFR genotype and identify patients with an EGFR mutation in a cost-effective and non-invasive manner.

Funding

This work was supported by a grant provided by Conquer Cancer Foundation of ASCO [2021IIG-5555960128] and Pfizer Products India Pvt. Ltd.

Vaxign-DL: A Deep Learning-based Method for Vaccine Design and its Evaluation

Reverse vaccinology (RV) provides a systematic approach to identifying potential vaccine candidates based on protein sequences. The integration of machine learning (ML) into this process has greatly enhanced our ability to predict viable vaccine candidates from these sequences. We have previously developed a Vaxign-ML program based on the eXtreme Gradient Boosting (XGBoost). In this study, we further extend our work to develop a Vaxign-DL program based on deep learning techniques. Deep neural networks assemble non-linear models and learn multilevel abstraction of data using hierarchically structured layers, offering a data-driven approach in computational design models. Vaxign-DL uses a three-layer fully connected neural network model. Using the same bacterial vaccine candidate training data as used in Vaxign-ML development, Vaxign-DL was able to achieve an Area Under the Receiver Operating Characteristic of 0.94, specificity of 0.99, sensitivity of 0.74, and accuracy of 0.96. Using the Leave-One-Pathogen-Out Validation (LOPOV) method, Vaxign-DL was able to predict vaccine candidates for 10 pathogens. Our benchmark study shows that Vaxign-DL achieved comparable results with Vaxign-ML in most cases, and our method outperforms Vaxi-DL in the accurate prediction of bacterial protective antigens.

Development of chimeric protein as a multivalent vaccine for human Kinetoplastid infections: Chagas disease and leishmaniasis

Leishmania spp. and Trypanosoma cruzi are parasitic kinetoplastids of great medical and epidemiological importance since they are responsible for thousands of deaths and disability-adjusted life-years annually, especially in low- and middle-income countries. Despite efforts to minimize their impact, current prevention measures have failed to fully control their spread. There are still no vaccines available. Taking into account the genetic similarity within the Class Kinetoplastida, we selected CD8+ T cell epitopes preserved among Leishmania spp. and T. cruzi to construct a multivalent and broad-spectrum chimeric polyprotein vaccine. In addition to inducing specific IgG production, immunization with the vaccine was able to significantly reduce parasite burden in the colon, liver and skin lesions from T. cruzi, L. infantum and L. mexicana challenged mice, respectively. These findings were supported by histopathological analysis, which revealed decreased inflammation in the colon, a reduced number of degenerated hepatocytes and an increased proliferation of connective tissue in the skin lesions of the corresponding T. cruzi, L. infantum and L. mexicana vaccinated and challenged mice. Collectively, our results support the protective effect of a polyprotein vaccine approach and further studies will elucidate the immune profile associated with this protection. Noteworthy, our results act as conceptual proof that a single multi-kinetoplastida vaccine can be used effectively to control different infectious etiologies, which in turn can have a profound impact on the development of a new generation of vaccines.

A Data-Driven Approach to Construct a Molecular Map of Trypanosoma cruzi to Identify Drugs and Vaccine Targets

Chagas disease (CD) is endemic in large parts of Central and South America, as well as in Texas and the southern regions of the United States. Successful parasites, such as the causative agent of CD, Trypanosoma cruzi have adapted to specific hosts during their phylogenesis. In this work, we have assembled an interactive network of the complex relations that occur between molecules within T. cruzi. An expert curation strategy was combined with a text-mining approach to screen 10,234 full-length research articles and over 200,000 abstracts relevant to T. cruzi. We obtained a scale-free network consisting of 1055 nodes and 874 edges, and composed of 838 proteins, 43 genes, 20 complexes, 9 RNAs, 36 simple molecules, 81 phenotypes, and 37 known pharmaceuticals. Further, we deployed an automated docking pipeline to conduct large-scale docking studies involving several thousand drugs and potential targets to identify network-based binding propensities. These experiments have revealed that the existing FDA-approved drugs benznidazole (Bz) and nifurtimox (Nf) show comparatively high binding energies to the T. cruzi network proteins (e.g., PIF1 helicase-like protein, trans-sialidase), when compared with control datasets consisting of proteins from other pathogens. We envisage this work to be of value to those interested in finding new vaccines for CD, as well as drugs against the T. cruzi parasite.

Vaxi-DL: An Artificial Intelligence-Enabled Platform for Vaccine Development

Vaccine development is a complex and long process. It involves several steps, including computational studies, experimental analyses, animal model system studies, and clinical trials. This process can be accelerated by using in silico antigen screening to identify potential vaccine candidates. In this chapter, we describe a deep learning-based technique which utilizes 18 biological and 9154 physicochemical properties of proteins for finding potential vaccine candidates. Using this technique, a new web-based system, named Vaxi-DL, was developed which helped in finding new vaccine candidates from bacteria, protozoa, viruses, and fungi. Vaxi-DL is available at: https://vac.kamalrawal.in/vaxidl/ .

COVID-19 Vaccines: Computational tools and Development

The 2019 coronavirus outbreak, also known as COVID-19, poses a serious threat to global health and has already had widespread, devastating effects around the world. Scientists have been working tirelessly to develop vaccines to stop the virus from spreading as much as possible, as its cure has not yet been found. As of December 2022, 651,918,402 cases and 6,656,601 deaths had been reported. Globally, over 13 billion doses of vaccine have been administered, representing 64.45% of the world's population that has received the vaccine. To expedite the vaccine development process, computational tools have been utilized. This paper aims to analyze some computational tools that aid vaccine development by presenting positive evidence for proving the efficacy of these vaccines to suppress the spread of the virus and for the use of computational tools in the development of vaccines for emerging diseases.

Identification of vaccine targets & design of vaccine against SARS-CoV-2 coronavirus using computational and deep learning-based approaches

An unusual pneumonia infection, named COVID-19, was reported on December 2019 in China. It was reported to be caused by a novel coronavirus which has infected approximately 220 million people worldwide with a death toll of 4.5 million as of September 2021. This study is focused on finding potential vaccine candidates and designing an in-silico subunit multi-epitope vaccine candidates using a unique computational pipeline, integrating reverse vaccinology, molecular docking and simulation methods. A protein named spike protein of SARS-CoV-2 with the GenBank ID QHD43416.1 was shortlisted as a potential vaccine candidate and was examined for presence of B-cell and T-cell epitopes. We also investigated antigenicity and interaction with distinct polymorphic alleles of the epitopes. High ranking epitopes such as DLCFTNVY (B cell epitope), KIADYNKL (MHC Class-I) and VKNKCVNFN (MHC class-II) were shortlisted for subsequent analysis. Digestion analysis verified the safety and stability of the shortlisted peptides. Docking study reported a strong binding of proposed peptides with HLA-A*02 and HLA-B7 alleles. We used standard methods to construct vaccine model and this construct was evaluated further for its antigenicity, physicochemical properties, 2D and 3D structure prediction and validation. Further, molecular docking followed by molecular dynamics simulation was performed to evaluate the binding affinity and stability of TLR-4 and vaccine complex. Finally, the vaccine construct was reverse transcribed and adapted for E. coli strain K 12 prior to the insertion within the pET-28-a (+) vector for determining translational and microbial expression followed by conservancy analysis. Also, six multi-epitope subunit vaccines were constructed using different strategies containing immunogenic epitopes, appropriate adjuvants and linker sequences. We propose that our vaccine constructs can be used for downstream investigations using in-vitro and in-vivo studies to design effective and safe vaccine against different strains of COVID-19.