Chronic liver disease is a major and growing health and economic burden on society. Liver disease is usually asymptomatic until obvious complications and severe liver damage occur. Given the progressive and generally unrecognized course of liver disease, it is important to apply modern "omics" technologies to identify specific proteins that may play a causal role in the disease. By linking genetic factors to disease via specific proteins, it is possible to highlight potential therapeutic targets, tailor existing drugs to new indications for the disease, and reduce potential doubts about the safety of drugs in development. Classification of specific proteins is essential for processing the results of proteomic analyzes. In this work, the classification system PANTHER and the database STRING were used to predict protein interactions. The classification system PANTHER (Protein ANalysis THrough Evolutionary Relationships) is used to classify proteins (and their genes) to facilitate high throughput analysis. STRING is a database of known and computer-calculated protein-protein interactions compiled from interactions in other (primary) databases. The results of proteomic analysis of published papers in the field of hepatotoxicity were used for data processing. The programs used sorted the proteins according to their functions, the process they are involved in, their position and pathway in the cell, and classified them into direct and indirect functional relationships. This allows the processing of the results of the proteome analysis through an in-depth analysis of the totality of its proteins, the liver proteome.