Finding Potential RNA Aptamers for a Protein Target Using Sequence and Structure Features

초록

Aptamers are single-stranded DNA or RNA sequences that tightly bind to a specific target molecule. This paper presents a computational method for generating potential protein-binding RNA aptamers using a random forest based on several features of protein and RNA sequences and RNA secondary structures. The results of cross validation and independent testing showed that our method can significantly reduce the initial pool of RNA sequences and that the top 10 candidates of RNA aptamers have similar secondary structures and protein-binding structures as actual RNA aptamers for a protein target. Although preliminary, our approach will be useful for constructing an initial pool of RNA sequences for experimental selection of aptamers.