GATCGGenerator: New Software for Generation of Quasirandom Nucleotide Sequences

Abstract

Introduction. In  recent  decades,  knowledge  about  DNA  has  been  increasingly  used  to  solve  biological  problems (calculations using DNA, long-term storage of information). Principally, we are talking about cases when it is required to select artificial nucleotide sequences. Special programs are used to create them. However, existing generators do not take into account the physicochemical properties of DNA and do not allow obtaining sequences with a pronounced “non-biological” structure. In fact, they generate sequences by distributing nucleotides randomly. The objective of this work is to create a generator of quasirandom sequences with a special nucleotide structure. It should take into account some physicochemical features of nucleotide structures, and it will be involved in storing non-biological information in DNA.Materials and Methods. A new GATCGGenerator software for generating quasirandom sequences of nucleotides was described. It was presented as SaaS (from “software as a service”), which provided its availability from various devices and platforms. The program generated sequences of a certain structure taking into account the guanine-cytosine (GC) composition and the content of dinucleotides. The performance of the new program algorithm was presented. The requirements for the generated nucleotide sequences were set using a chat in Telegram, the interaction with the user was clearly shown. The differences between the input parameters and the specific nucleotide structures obtained as a result of the program were determined and generalized. Also, the time costs of generating sequences for different input data were given in comparison. Short sequences differing in type, length, GC composition and dinucleotide content were studied. The tabular form shows how the input and output parameters are correlated in this case.Results. The developed software was compared to existing nucleotide sequence generators. It has been established that the generated sequences differ in structure from the known DNA sequences of living organisms, which means that they can be used as auxiliary or masking oligonucleotides suitable for molecular biological manipulations (e.g., amplification reactions), as well as for storing non-biological information (images, texts, etc.) in DNA molecules. The proposed solution makes it possible to form specific sequences from 20 to 5 000 nucleotides long with a given number of dinucleotides and without homopolymer fragments. More stringent generation conditions remove known limitations and provide the creation of quasirandom sequences of nucleotides according to specified input parameters. In addition to the number and length of sequences, it is possible to determine the GC composition, the content of dinucleotides, and the nature of the nucleic  acid  (DNA  or  RNA)  in  advance.  Examples  of  short  sequences  differing  in  length,  GC  composition  and dinucleotide content are given. The obtained 30-nucleotide sequences were tested. The absence of 100 % homology with known DNA sequences of living organisms was established. The maximum coincidence was observed for the generated sequences with a length of 25 nucleotides (similarity of about 80 %). Thus, it has been proved that GATCGGenerator can generate non-biological nucleotide sequences with high efficiency.Discussion and Conclusion. The new generator provides the creation of nucleotide sequences in silico with a given GC composition. The solution makes it possible to exclude homopolymer fragments, which improves qualitatively the physicochemical stability of sequences.