Mercury

also available in : 한국어 / 日本語

Mercury is an open source translator designed to transform human-readable natural language to DNA Sequences. The project aims for a unique bridge between complex genomic information and accessible linguistic representations, making biological data more intuitive for researchers, developers, and enthusiasts.

The core functionality of mercury is to analyze DNA sequences and convert them into meaningful natural language sentences or designated binary sequences. This allows for DNA-embedded messages to be translated into human-readable format.

Disclaimer

The current version of mercury does not guarantee the stability of DNA sequences when used in any types of real life forms, in that many of the problems including GC ratio, pre-designated sequences for ribosomes, etc. are yet to be solved. Version 2 may include features to handle limitations mentioned above, but the version update cannot be promised at the moment.

Features

• Bidirectional Conversion:

  • DNA → Natural Language / Binary
  • Natural Language / Binary → DNA

• Codon-Aware Encoding:

  • All DNA sequences include start/stop codons and are padded to maintain 3-base alignment

DNA Encoding Background

DNA is composed of four nucleotide bases: A, C, G, and T. Mercury leverages this limited but information-rich alphabet to encode binary or textual data, aligning with biological conventions such as:

• Start Codon: ATG (initiation signal)
• Stop Codons: TAA, TAG, TGA (termination signal)
• Codon Alignment: DNA is interpreted in triplets (3 bases)

Mercury ensures all generated DNA sequences follow this codon structure for biological validity and decoding accuracy.

Conversion Process

Conversion Table

Mercury uses a 2-bit representation for each nucleotide base, allowing DNA sequences to encode binary data and vice versa. Below is the default mapping:

Nucleotide	Binary
A	00
C	01
G	10
T	11

This table is used when converting binary sequences to DNA and decoding DNA back to binary data.

Padding

DNA sequences must be aligned in triplets (codons) to preserve biological structure.
Since Mercury uses 2-bit per base encoding (4 bases per byte), the resulting DNA sequence length may not always be divisible by 3.

To ensure correct codon alignment, Mercury performs padding based on the following rules:

1. Padding Bases (A)

If the payload (converted to DNA) is not a multiple of 3 in length, Mercury appends padding bases (A, representing 00) at the end of the binary-encoded DNA sequence.
These padding bases ensure the total DNA length before adding the stop codon is divisible by 3.

2. Padding Length Determination & Stop Codon Selection

The number of padding bases added determines which stop codon is used.
This mechanism allows the decoder to infer the original payload length by examining the stop codon, effectively eliminating the need for an explicit length header.

Stop Codon	Padding Length (bases)
TAA	0 (no padding)
TAG	1 base padding
TGA	2 bases padding

---

Example

Input

• Text: "A"
• Encoding: UTF-8 → 0x41 → Binary: 01000001 (8 bits)

Process

1. Binary → DNA Payload (2 bits per base):

Bits	DNA base
01	C
00	A
00	A
01	C

Resulting DNA Payload: CAAC (4 bases)

2. Determine Padding: Payload length = 4 bases
   4 % 3 = 1 → length is 3n + 1 According to the padding rule, a remainder of 1 requires 2 bases of padding.

3. Apply Padding: Append 2 padding bases (A) to DNA payload:
   CAAC + AA = CAACAA
   Now length = 6 bases (divisible by 3)

4. Select Stop Codon: Since 2 padding bases were added, select stop codon: TGA

---

Final DNA sequences: ATGCAACAATGA

How to Use

Gradle

dependencies {
    implementation("io.github.hodadako:mercury-core:0.1.2")
}

In Code

MercuryTranslator translator = new SimpleMercuryTranslator();

String dna = translator.encode("I will be always attached to you");
// ATGCAGCAGAACTCTCGGCCGTACGTAAGAACGAGCGCCAGAACGACCGTACTCTCGACCTGCCTATAGAACGACCTCACTCACGACCGATCGGACGCCCGCAAGAACTCACGTTAGAACTGCCGTTCTCCATAG
String text = translator.decode(dna);
// I will be always attached to you

How to Contribute

TBD

Issues & Bugs

If you encounter any issues or bugs while using Mercury, please feel free to open an issue on the GitHub Issues page.

When reporting a bug, kindly include the following information to help us diagnose and resolve the problem more efficiently:

• Steps to reproduce the issue
• Expected vs. actual behavior
• Your environment details (OS, Java version, Mercury version)
• Any relevant error messages or stack traces

We appreciate your feedback and contributions to improve Mercury!

Inspired By

This project was mainly inspired by the great passion of the original author to the G-Witch series.

Spoiler Alert: Key Scene Description

In a memorable scene, the protagonist Miorine Rembran discovers a hidden message encoded in the DNA of a tomato left by her mother.

The anime beautifully depicts this moment where a security system is unlocked using a Java-based system designed to decode the DNA message—mirroring the core idea behind this project.

Blessing.