Method and apparatus for predicting enzyme commission number using amino acid sequence

A non-transitory computer-readable medium or media comprising one or more sequences of instructions which, when executed by a processor, causes steps for predicting an enzyme commission number using an amino acid sequence of a subject, comprising:

• receiving data of the amino acid sequence
• preprocessing including converting the data of the amino acid sequence into a 2-dimensional embedding matrix having M rows and N columns, creating newly a plurality of the 2-dimensional embedding matrices by shifting at least one row of M rows in a column direction of the 2-dimensional embedding matrix, and transforming the plurality of the 2-dimensional embedding matrices into a 3-dimensional embedding matrix by stacking the plurality of the 2-dimensional embedding matrices together;
• extracting a feature of the amino acid sequence from the 3-dimensional embedding matrix using a machine learning model;
• and
• predicting the enzyme commission number based on the extracted feature of the amino acid sequence.