Abstract To improve the ergonomic qualities of graphical interfaces and optimize the visual perception of large data presented in text form, a method of information processing based on DNA algorithms is proposed. These algorithms were developed by the author in the study of the molecular genetic and structural properties of long nucleotide sequences. The result of processing information presented in text form is an object of discrete geometry, two-dimensional projections of which can serve for visual identification of the internal structure of information, as well as for interpretation as characteristic symbols. At the same time, the considered transformations have the property of visualizing the characteristic cluster structure of information. It is shown that genetic nucleotide sequences, as well as texts in natural language, lend themselves to visual classification, which contributes to an increase in the productivity of mental work when analyzing big data. It is shown that DNA algorithms can reduce the load on the visual analyzer during the initial examination of data presented in text form. This makes it possible to apply this method when operators interact with various information streams and text data of arbitrary nature, as well as to improve the ergonomic qualities of specialized graphical interfaces.