AI assisted encryption into DNA sequence of a functional protein

Abstract

For thousands of years, a war has been waged between hackers and cryptographers1. Cryptographers find a safe way to store and transport information, and hackers attempt to access it. Today, this war is waged on a much more microscopic scale, and as new storage methods, like ones utilizing DNA are developed, cryptographers need to find a way to protect the data from malicious entities1,2. A solution may lie in the thousands of proteins in the human cell and attaching information onto these proteins while preserving their function. Even though there are encryption schemes that managed to insert information onto DNA a reliable approach to consistently preserve the function of the protein, and hence the viability of the cells transporting the information is necessary2,3. As a result, the motivation of this study was to devise a way to efficiently hide a message in living cells that cannot be discovered without DNA sequencing. We utilized these proteins, coupled with this Artificial Intelligence (AI) centered approach, to devise a standard scheme where one could reliably encode encrypted information onto these functional proteins. On the basis of repeated predictability modeling, and wet lab generations of the model system, Enhance Green Fluorescent Protein (EGFP). Our group attempted to develop a program capable of reliably encrypting information onto a protein of interest, EGFP, based on a desired degree of functionality and the amount of information needed to be encrypted. The encryption done was based on the Advanced Encryption standard (AES), the golden standard of encryption established by the U.S. National Institute of Standards and Technology (NIST)4. This scheme, similar to the work done centuries ago, contributes to the added security and applicability of novel data storage and information transfer methods in the field of synthetic biology.