VacSol: a high throughput in silico pipeline to predict potential therapeutic targets in prokaryotic pathogens using subtractive reverse vaccinology
- Muhammad Rizwan†1,
- Anam Naz†2,
- Jamil Ahmad1Email author,
- Kanwal Naz2,
- Ayesha Obaid2,
- Tamsila Parveen3,
- Muhammad Ahsan1 and
- Amjad Ali2Email author
© The Author(s). 2017
Received: 27 September 2016
Accepted: 8 February 2017
Published: 13 February 2017
With advances in reverse vaccinology approaches, a progressive improvement has been observed in the prediction of putative vaccine candidates. Reverse vaccinology has changed the way of discovery and provides a mean to propose target identification in reduced time and labour. In this regard, high throughput genomic sequencing technologies and supporting bioinformatics tools have greatly facilitated the prompt analysis of pathogens, where various predicted candidates have been found effective against certain infections and diseases. A pipeline, VacSol, is designed here based on a similar approach to predict putative vaccine candidates both rapidly and efficiently.
VacSol, a new pipeline introduced here, is a highly scalable, multi-mode, and configurable software designed to automate the high throughput in silico vaccine candidate prediction process for the identification of putative vaccine candidates against the proteome of bacterial pathogens. Vaccine candidates are screened using integrated, well-known and robust algorithms/tools for proteome analysis, and the results from the VacSol software are presented in five different formats by taking proteome sequence as input in FASTA file format. The utility of VacSol is tested and compared with published data and using the Helicobacter pylori 26695 reference strain as a benchmark.
VacSol rapidly and efficiently screens the whole bacterial pathogen proteome to identify a few predicted putative vaccine candidate proteins. This pipeline has the potential to save computational costs and time by efficiently reducing false positive candidate hits. VacSol results do not depend on any universal set of rules and may vary based on the provided input. It is freely available to download from: https://sourceforge.net/projects/vacsol/.
KeywordsReverse vaccinology Computational pipeline Vaccine candidates Subtractive proteomics PVCs VacSol
In silico prediction of vaccine candidates has great significance in various life science disciplines, including biomedical research . The conventional approach of vaccine development requires pathogenic cultivation in vitro that is not always possible. Although this methodology has the potential to produce successful vaccines and has long been in practice, but now considered time-consuming and inadequate for most pathogens. This caveat is particularly evident when microbes are inactive, protective, or even in the case where antigen expression is decreased; rendering the conventional approach a significant challenge for putative vaccine candidate discovery [2, 3]. These basic problems have led scientists to develop new vaccinology approaches based on advanced computational tools. In particular, with the introduction of high-throughput sequencing techniques over the last decade and the advent of bioinformatics approaches, Rino Rappouli revolutionized Pasteur’s vaccinology procedure by introducing a novel “reverse vaccinology” method [4–6]. This advanced in-silico technique for vaccine prediction couples genomic information and analysis with bioinformatics tools. Using this approach, several vaccines have been successfully developed against microbial pathogens [7–9]. Reverse vaccinology is now recognized as safer and more reliable as compared to conventional vaccinology methods [10, 11].
Using the reverse vaccinology approach, various predictive and analytical tools (Vaxign, VaxiJen, JennerPredict) have been designed for the identification of putative vaccine candidates. These tools are widely available online [12–14], but only a handful of softwares and pipelines, like NERVE and Vacceed [15, 16], are accessible as full packages. Although web-based pipelines are efficient, their drawbacks include time delays and constraints for input file size.
NERVE (New Enhanced Reverse Vaccinology Environment), a Perl based modular pipeline for in-silico identification of potential vaccine candidates, generates results through text interface configuration and is an efficient, modular-based standalone software for vaccine candidate identification . But it only focuses on adhesion proteins whereas several non-adhesion proteins can also participate in host-pathogen interactions (including porin, flagellin, invasin, etc.), and most of them are pathogenic as well as antigenic. Therefore, there exists a perilous need for an updated and advanced analysis tool that inclusively provides every putative candidate in its output.
Vacceed is another highly configurable architecture designed to perform high throughput in silico identification of eukaryotic vaccine candidates. Vacceed is, in fact, able to reduce false vaccine candidates that are selected for laboratory validation to save time and money , but this highly efficient, scalable, and configurable program provides limited information on pathogenicity and putative functional genes. These main parameters prove instrumental in the determination of potential vaccine candidates. Thus, given the current software limitations, we sought to utilize the reverse vaccinology approach to overcome limitations of currently available pipelines.
We therefore focused on in silico reverse vaccinology approach to address the issues that were present in previous pipelines, and to precisely screen out the putative vaccine candidates from whole bacterial genome in silico. We designed a new automated pipeline, termed VacSol, to efficiently screen for the therapeutic vaccine agents from the bacterial pathogen proteome to save both time and resources.
Tools and databases integrated and implemented in VacSol
New command line sequence alignment application developed using the NCBI C++ toolkit.
Package of programs that support the search method of generalized profile formatting.
Protein subcellular localization prediction tool.
Transmembrane topology prediction tool.
Database of essential genes.
Virulence factors database.
B-Cell epitope prediction tool.
Prediction of promiscuous major histocompatibility complex (MHC) Class-I binding sites.
Prediction of MHC Class-II binding regions in an antigen sequence.
Manually annotated protein sequences database with information extracted from literature.
VacSol performs various proteome-wide analyses and generates results in five different formats. This pipeline was validated using a sample data set of the Helicobacter pylori proteome. The selected strain of H. pylori 26695 (RefSeq NC_000915.1) is comprised of 1576 proteins or coding regions , and the whole proteome was scanned in each protein prioritizing step.
Implementation of VacSol for test data
The first working step was performed by identifying the non-host homologs, required to elute host homologous proteins to restrict the chance of autoimmunity [20, 21]. Out of 1576 possible proteins, 1452 were screened as non-human homologous proteins by using BLASTp against RefSeq  and SwissProt  databases. For BLAST non-human homologs, criteria included a Bit Score >100, E-Value <1.0 e(−5), and percentage identity >35% . Next, these 1452 proteins were subjected for further protein prioritization processing by VacSol to predict subcellular localization. 65 proteins were found to be in the secretome and exoproteome, of which 23 proteins lie in the extracellular region, and 42 were screened as outer-membrane proteins. Prioritization of proteins according to localization substantially contributed to enhance the PVCs identification process . Surface exposed proteins tend to be involved in pathogenesis, making them prime targets as vaccine candidates . Similarly, both extracellular and secreted proteins are readily accessible to antibodies as compared to intracellular proteins, and therefore represent ideal vaccine candidates. Results obtained through PSORTb, and integrated in VacSol, were then cross-checked with CELLO2GO  to confirm the localization of putative candidate proteins. After localization validation, screened proteins were checked for their essentiality. 667 proteins were sorted as essential genes required for the survival of gastric pathogen H. pylori. Finally, 10 proteins have been prioritized following all the criteria. This analysis reduced the cost and time of PVCs identification by excluding proteins with no suitable features for further processing.
The Database of Essential Genes (DEG)  was then used to predict essential genes. Results demonstrated that all 10 of the prioritized proteins were essential proteins, thus making them putative vaccine candidates. In the next step, the proteome was screened for virulent proteins, as identification of virulent factors in essential proteins is a key step in the vaccine development process . Essential genes of a pathogen tend to be virulent, substantiating these checks as key factors in the prediction of target proteins to prioritize vaccine candidates [21, 30]. In our case, 267 proteins were found to be virulent proteins among whole proteome of the pathogen.
Functional annotation of prioritized proteins
Protein ID (VacSol)
Gene symbol (NCBI)
Molecular weight kDa (ExPASy)
Molecular function (UNIPROT)
Domains (Interpro Scan)
Iron(III) dicitrate transport protein (FecA)
TonB-dependent receptor & plug domain
Flagellin A (FlaA)
Cell motility, Signal transduction and structural molecule activity
Flagellin, Flagellin_D0/D1, Flagellin_hook_IN_motif
Sel1-like, TPRlike_ helical_dom, TPR_2
Iron(III) dicitrate transport protein (FecA)
TonB-dependent receptor & plug domain
Flagellin B (FlaB)
Structural molecule activity
Toxin-like outer membrane protein
Autotransport_beta& Vacuolating_cytot oxin_put
Toxin-like outer membrane protein
VacA2 (motif), Autotransporte_beta, PbH1
Peptidoglycan, cell wall synthesis
Toxin-like outer membrane protein
Vacuolating cytotoxin putative & Autotransporter beta domain
Iron(III) dicitrate transport protein (FecA)
TonB-dependent receptor, betabarrel, plug domain
The prioritized putative vaccine targets against H. pylori 26695 included FecA (HP1400), FecA (HP0807), FecA (HP0686), FlaA (HP0601), FlaB (HP0115), HcpA (HP0211), HcpC (HP1098), and toxin-like outer membrane proteins (HP0289, HP0610, and HP0922). Among these target candidates, Iron (III) dicitrate transport protein, FecA (HP1400, HP0807, and HP0686), interacts with TonB, a protein involved in the virulence process. Previous studies have shown that controlled and mutated TonB leads to increased immunization . Indeed, by targeting HP1400, HP0807, and HP0686, TonB can be controlled, making these three promising putative vaccine candidates.
Flagelline proteins (flaA and flaB) are responsible for the pro-inflammation of gastric mucosa that leads to the development of gastric/peptic ulcers, making flaA and flaB considerable candidates for novel vaccine development . Likewise, Beta-lactamase HcpA and HcpC are highly pathogenic proteins that are directly involved in different infections caused by H. pylori . The HcpA protein is also involved in bacterial and eukaryotic host interaction . These protein annotations verify that VacSol limited its screening to the proteins that are biologically relevant putative and therapeutic vaccine candidates.
Previous studies have linked three toxin-like proteins with virulent proteins and vaccine candidates BabA, CagS, Cag6, HpaA, and VacA . Indeed, Cag proteins are also well-known pathogenic proteins, involved in pathogenic pathways, while the HcpA protein has been shown to be involved in bacterial and eukaryotic host interactions . Using our computational approach, we have designed the VacSol pipeline to further the field of vaccinology by reducing time, cost and trial burdens in novel putative vaccine candidate protein identification. Proteins predicted using this pipeline against H. pylori strain may serve as promising PVCs against gastric pathogens, as substantiated by previous findings in the literature. Further evaluation of these PVCs can lead to the development of novel and effective vaccines against H. pylori.
VacSol is a new, highly efficient, and user-friendly pipeline established for biological scientists, including those with limited expertise in computational analyses. VacSol has restricted the pool of promising PVCs from the whole bacterial pathogen proteome by automatizing the in silico reverse vaccinology approach for the prediction of highly antigenic targeted proteins, via a user controlled step-wise process. This new pipeline is an efficient tool in the contexts of time and computational/experimental costs by eliminating false positive candidates from laboratory evaluation. The modular structure of VacSol improves the prediction process of vaccine candidates with additional potential for future development in this field.
Availability and requirements
Project name: VacSol: An in silico pipeline to predict potential therapeutic targets
Project home page: https://sourceforge.net/projects/vacsol/files/
Archived version: Not available
Operating system(s): Linux
Programming language: Java
Other requirements (Pre Requisite Tools/Languages):
• PSORTb 
• NCBI BLAST+ 
• Pftools 
• Hmmtop 
• ABCPred 
• ProPred-I 
• ProPred 
Database of Essential Genes
Iron (III) dicitrate transport protein A
Flagelline protein A
Flagelline protein B
- H. pylori :
Helicobacter cysteine-rich protein A
Helicobacter cysteine-rich protein C
Potential vaccine candidates
Virulence factor database
We acknowledge Andreana N. Holowatyj (Ph.D) from Department of Biological Sciences, Wayne State University School of Medicine, USA for proofreading the manuscript.
• Any restrictions to use by non-academics: No
No funding was provided for this project.
Availability of data and materials
VacSol is tested on Ubuntu and can be freely downloaded from:
The VacSol Installation and User Guide can be obtained from:
https://sourceforge.net/projects/vacsol/files/Installation and User Guide.docx/download/
H. pylori 26695 dataset used for analysis:
Protein sequences and their locations:
H. pylori 26695 full genome:
Full genome was retrieved from NCBI RefSeq with reference number NC_000915.1, available at following link.
AA conceived the idea. MR, JA, AN and AA designed the pipeline. MR implemented the software. AN and MR contributed to software validation. AN and MR composed the manuscript. JA, KN, AO, TP, and MA contributed to analyses and results, as well as in the drafting of the manuscript. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Tarca AL, et al. Machine learning and its applications to biology. PLoS Comput Biol. 2007;3(6):e116.View ArticlePubMedPubMed CentralGoogle Scholar
- Capecchi B, et al. The genome revolution in vaccine research. Curr Issues Mol Biol. 2004;6:17–28.PubMedGoogle Scholar
- Kimman T. Risks connected with the use of conventional and genetically engineerd vaccines. Vet Q. 1992;14(3):110–8.View ArticlePubMedGoogle Scholar
- Rinaudo CD, et al. Vaccinology in the genome era. J Clin Invest. 2009;119(9):2515–25.View ArticlePubMedPubMed CentralGoogle Scholar
- Pizza M, et al. Identification of vaccine candidates against serogroup B meningococcus by whole-genome sequencing. Science. 2000;287(5459):1816–20.View ArticlePubMedGoogle Scholar
- Rappuoli R. Reverse vaccinology, a genome-based approach to vaccine development. Vaccine. 2001;19(17):2688–91.View ArticlePubMedGoogle Scholar
- Ariel N, et al. Search for potential vaccine candidate open reading frames in the Bacillus anthracis virulence plasmid pXO1: in silico and in vitro screening. Infect Immun. 2002;70(12):6817–27.View ArticlePubMedPubMed CentralGoogle Scholar
- Montigiani S, et al. Genomic approach for analysis of surface proteins in Chlamydia pneumoniae. Infect Immun. 2002;70(1):368–79.View ArticlePubMedPubMed CentralGoogle Scholar
- Ross BC, et al. Identification of vaccine candidate antigens from a genomic analysis of Porphyromonas gingivalis. Vaccine. 2001;19(30):4135–42.View ArticlePubMedGoogle Scholar
- Rappuoli R. Reverse vaccinology. Curr Opin Microbiol. 2000;3(5):445–50.View ArticlePubMedGoogle Scholar
- Hassan A, et al. Pangenome and immuno-proteomics analysis of Acinetobacter baumannii strains revealed the core peptide vaccine targets. BMC Genomics. 2016;17(1):732.View ArticlePubMedPubMed CentralGoogle Scholar
- He Y, Xiang Z, Mobley HL. Vaxign: the first web-based vaccine design program for reverse vaccinology and applications for vaccine development. Biomed Res Int. 2010;2010:29750.Google Scholar
- Doytchinova IA, Flower DR. VaxiJen: a server for prediction of protective antigens, tumour antigens and subunit vaccines. BMC Bioinformatics. 2007;8(1):4.View ArticlePubMedPubMed CentralGoogle Scholar
- Jaiswal V, et al. Jenner-predict server: prediction of protein vaccine candidates (PVCs) in bacteria based on host-pathogen interactions. BMC Bioinformatics. 2013;14(1):211.View ArticlePubMedPubMed CentralGoogle Scholar
- Vivona S, Bernante F, Filippini F. NERVE: new enhanced reverse vaccinology environment. BMC Biotechnol. 2006;6(1):35.View ArticlePubMedPubMed CentralGoogle Scholar
- Goodswen SJ, Kennedy PJ, Ellis JT. Vacceed: a high-throughput in silico vaccine candidate discovery pipeline for eukaryotic pathogens based on reverse vaccinology. Bioinformatics. 2014;30:2381–3.View ArticlePubMedPubMed CentralGoogle Scholar
- Gardy JL, et al. PSORTb v. 2.0: expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis. Bioinformatics. 2005;21(5):617–23.View ArticlePubMedGoogle Scholar
- Petty DR, Sylvester SA. Specification language for defining user interface panels that are platform-independent. 2002. Google Patents.Google Scholar
- Baltrus DA, et al. The complete genome sequence of Helicobacter pylori strain G27. J Bacteriol. 2009;191(1):447–8.View ArticlePubMedGoogle Scholar
- Kotb M. Infection and autoimmunity: a story of the host, the pathogen, and the copathogen. Clin Immunol Immunopathol. 1995;74(1):10–22.View ArticlePubMedGoogle Scholar
- Naz A, et al. Identification of putative vaccine candidates against Helicobacter pylori exploiting exoproteome and secretome: a reverse vaccinology based approach. Infect Genet Evol. 2015;32:280–91.View ArticlePubMedGoogle Scholar
- Pruitt KD, Tatusova T, Maglott DR. NCBI reference sequences (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res. 2007;35 suppl 1:D61–5.View ArticlePubMedGoogle Scholar
- Boeckmann B, et al. The SWISS-PROT protein knowledgebase and its supplement TrEMBL in 2003. Nucleic Acids Res. 2003;31(1):365–70.View ArticlePubMedPubMed CentralGoogle Scholar
- Pertsemlidis A, Fondon JW, John W. Having a BLAST with bioinformatics (and avoiding BLASTphemy). Genome Biol. 2001;2(10):1–10.View ArticleGoogle Scholar
- Kaufmann SH, Lambert PH. The grand challenge for the future: vaccines for poverty-related diseases from bench to field. Springer Science & Business Media; 2005. http://www.springer.com/br/book/9783764371753.
- Grandi G. Bacterial surface proteins and vaccines. Ital J Biochem. 2007;56(3):R.Google Scholar
- Yu C-S, et al. CELLO2GO: a Web server for protein subCELlular LOcalization prediction with functional gene ontology annotation. PLoS One. 2014;9(6):e99368.View ArticlePubMedPubMed CentralGoogle Scholar
- Luo H, et al. DEG 10, an update of the database of essential genes that includes both protein-coding genes and noncoding genomic elements. Nucleic Acids Res. 2013;42(D1):D574–D80.View ArticlePubMedPubMed CentralGoogle Scholar
- Handman E. Leishmaniasis: current status of vaccine development. Clin Microbiol Rev. 2001;14(2):229–43.View ArticlePubMedPubMed CentralGoogle Scholar
- Muhammad SA, et al. Prioritizing drug targets in Clostridium botulinum with a computational systems biology approach. Genomics. 2014;104(1):24–35.View ArticlePubMedGoogle Scholar
- Chen L, et al. VFDB: a reference database for bacterial virulence factors. Nucleic Acids Res. 2005;33 suppl 1:D325–8.PubMedGoogle Scholar
- Tusnady GE, Simon I. The HMMTOP transmembrane topology prediction server. Bioinformatics. 2001;17(9):849–50.View ArticlePubMedGoogle Scholar
- Consortium U. The universal protein resource (UniProt). Nucleic Acids Res. 2008;36 suppl 1:D190–5.Google Scholar
- Hsieh P-F, et al. Serum-induced iron-acquisition systems and TonB contribute to virulence in Klebsiella pneumoniae causing primary pyogenic liver abscess. J Infect Dis. 2008;197(12):1717–27.View ArticlePubMedGoogle Scholar
- Constantinescu C, Constantinescu E. Could knowledge of H. pylori pathogenicity factors lead to the emergence of new methods for identifying bacteria? Bull Transilvania University of Brasov, Seriels VI: Medical Sciences. 2014;7:1.Google Scholar
- Mittl PR, et al. Detection of high titers of antibody against Helicobacter cysteine-rich proteins A, B, C, and E in Helicobacter pylori-infected individuals. Clin Diagn Lab Immunol. 2003;10(4):542–5.PubMedPubMed CentralGoogle Scholar
- Mittl PR, Schneider-Brachert W. Sel1-like repeat proteins in signal transduction. Cell Signal. 2007;19(1):20–31.View ArticlePubMedGoogle Scholar
- Camacho C, et al. BLAST+: architecture and applications. BMC Bioinformatics. 2009;10(1):421.View ArticlePubMedPubMed CentralGoogle Scholar
- Sigrist CJ, et al. PROSITE: a documented database using patterns and profiles as motif descriptors. Brief Bioinform. 2002;3(3):265–74.View ArticlePubMedGoogle Scholar
- Saha S, Raghava G. Prediction of continuous B‐cell epitopes in an antigen using recurrent neural network. Proteins. 2006;65(1):40–8.View ArticlePubMedGoogle Scholar
- Singh H, Raghava G. ProPred1: prediction of promiscuous MHC Class-I binding sites. Bioinformatics. 2003;19(8):1009–14.View ArticlePubMedGoogle Scholar
- Singh H, Raghava G. ProPred: prediction of HLA-DR binding sites. Bioinformatics. 2001;17(12):1236–7.View ArticlePubMedGoogle Scholar