Determination of Mutation Pattern in Human Androgen Receptor by Means of Amino-Acid Pair Predictability

The mutation trend and pattern in protein are currently studied directly using amino acid sequence, however, it would be more efficient if the amino acid sequence is transferred into other domains through quantification because sophisticated mathematical tools can be applied to a numeric sequence. In this study, we apply the amino-acid pair predictability to quantifying the human androgen receptor with its 215 missense point mutations to analyze which amino-acid pairs are sensitive to mutations. The results show that 94.88% mutations occurred at the unpredictable amino-acid pairs, 79% mutations targeted at one or two original amino-acid pairs whose actual frequency is larger than predicted frequency and 63.26% mutations lead to one or two mutated amino-acid pairs with their actual frequency smaller than predicted one.