Figure 1.

Schematic explanation of one-hot encoding, zero-padding and truncation of amino acid sequences (A) Amino acid sequences of different lengths are shaped to the common dimension of 7 by truncating or padding zeros at the end. (B) Amino acid sequence at common length L is transformed to a binary matrix $(n+1)\times L$, being n the number of different amino acids and placeholders. Each column of this matrix is full of zeros, being one only in the position of the corresponding amino acid.