RESUMO
During the construction of a random peptide repertoire using degenerate models, unexpected amino acids or stop codons are almost unavoidable. To conquer this shortcoming, a new split-mix-split method of oligonucleotide synthesis was developed. A 13-amino acids random peptide library had been constructed by using this method. The sequencing results of 16 clones indicated that neither stop codon nor codon for cysteine appeared as designed. The occurrence rations of 19 amino acids were also calculated and no obvious amino acid bias had been observed. By using this method, the type and quantity of amino acid at certain position of a peptide could be controlled well, so this split-mix-split method, combined with degenerate could meet the needs of a high diversity random peptide library.