Development of a bioinformatic pipeline for genome-wide prioritization of serological diagnostic markers in pathogens

BRUNER M; RAMOA D; CARMONA SJ; AGÜERO F

Santa Fe

Congreso; XXVIII Reunión Anual de la Sociedad Argentina de Protozoología y Enfermedades Parasitarias y SIMPOSIO Internacional de Biología Celular y Molecular de la Enfermedad de Chagas; 2016

Sociedad Argentina de Protozoología y Enfermedades Parasitarias (SAP)

Background: B-cell epitopes are the parts ofantigen molecules recognized by antibodymolecules produced by the immune system.Knowledge of B-cell epitopes may be used in thedesign of vaccines and diagnostics tests. Basedon the availability of complete genomes ofpathogens,weproposedanimprovedcomputational strategy to prioritize potential newantigens (in the form of short peptides) at agenomic scale. To validate this tool, antigens forten complete proteomes were ranked, following apublished method (see Carmona SJ et al 2012).The results were validated with data from differentdatabases and known, curated antigens andepitopes found in the literature. By ROC curveanalysis we demonstrate the performance of themethod to enrich in interesting candidate antigensat the top of the ranking. Results: In this work wewill present the results of prioritizing candidatemarkers from the following pathogen proteomes ofBorrelia burgdorferi, Francisella tularensis,Brucella melitensis, Coxiella burnetti, Leptospirainterrogans, Mycobacterium tuberculosis, M.leprae, Plasmodium falciparum, Toxoplasmagondii and Trypanosoma cruzi. For this, the tooldecomposed the proteomes into short peptides (ofa user-defined length) and analyzed 11features/properties related to their antigenicpotential,predictedsubcellularlocation,immunogenic potential, and sequence similaritywith other species, amongst others. Optimizedfeature weightings were obtained for eachorganism that maximize AUC values for knownantigens. Conclusion: We improved a method ofprioritizing of antigens and prediction of linear B-cell epitopes based on different properties ofproteins, mainly those that can be assessed usingthe protein primary structure. An important novelty of the method is that includes features that arenew to the B-cell epitope prediction by theaddition of global protein properties. This tool canbe used to guide the design of peptidemicroarrays or other methodologies that requireusers to limiting the search space when lookingfor new antigens for serological diagnosis.