Introduction. Antibodies to nucleic acids (NAs) represent a significant subject of investigation in medicine, yet they have not gained wide dissemination due to the low immunogenicity of NAs and the high cross-reactivity of antibodies, which can lead to autoimmune diseases when using modern mRNA vaccines. Diagnostic antibodies for detecting RNA antigens hold considerable importance in various domains of knowledge.

Methods. A promising predictive model, "polyclonal antibody - potato pathogen," can aid in enhancing the specificity of polyclonal antibodies to NAs. In this study, the selected pathogens include PSTVd, PVX, and PVY. Using Python, the program can identify unique sequences within the genome of the specified pathogen—sequences of user-defined length (ribotopes) absent in the host genome, which can subsequently serve as targets for antibody generation. Additionally, the program identifies the most unique sequence among those found by breaking them into all possible sub-sequences of, for example, length 4 (the minimum length of an NA where antibodies can bind to it) and more nucleotides. The program analyzes these subsequences for their occurrence in the host genome. The sequences with the least number of occurrences of their subsequences in the host genome are assigned the status of the most unique and are displayed on the screen.

Results. Calculations showed that the minimum length of a sequence ensuring the uniqueness of a ribotope is 11. Several unique sequences of this length were found in the genomes of each pathogen. Based on these findings, in vivo experiments are planned on laboratory mice to generate antibodies and test their specificity using the IFA method. Thus, obtaining highly specific antibodies is achieved through the proper selection of the "target".

Conclusions. The obtained model can potentially be extrapolated to human viral pathogens and can subsequently find application in molecular diagnostics as well as in medicine for creating higher quality analogs of modern mRNA vaccines.