zlice / sna Goto Github PK

Proof of Concept DNA storage
======================

SNA is a DNA digital storage encoder/decoder for base2 (binary) to base4 (DNA)

4ROLL method is for encoding Digital information in Biological format

Use this format for encoding, and 'roll' through a table to change the binary to DNA order
00 = A   ->   00 = C ...
01 = C   ->   01 = G ...
10 = G   ->   10 = T ...
11 = T   ->   11 = A ...
1byte = 4bases

Uses hardcoded values. Focus of code is a PoC, not usability or arguments.
(Willing to increase functionality if proven useful)

4ROLL.c uses a hardcode file /tmp/test to encode data
data is printed to stdout

4ROLLdec.c uses a hardcoded file /tmp/test.dec to decode data
data is automatically written to /tmp/output
/tmp/output will be overwritten

gentoo live iso pulled from https://gentoo.org/downloads
MLK mp3 pulled from EBI https://www.ebi.ac.uk/
https://www.ebi.ac.uk/goldman-srv/DNA-storage/orig_files/

===========
COMPILE
===========

gcc -o 4ROLL 4ROLL.c
gcc -o 4ROLLdec 4ROLLdec.c

===========
RUNNING
===========

./4ROLL > /path/to/output.extension
./4ROLLdec (automatically outputs to /tmp/output, can change in code for windows or symlink for *nix based systems)

======================
GENTOO ISO RUN & STATS
======================

Gentoo live iso is about 1.4GB
Encoded into DNA is about 5.9GB

$ date; ./4ROLL > gentoo.dna;date
Sat Mar 17 09:00:45 MST 2018
Sat Mar 17 09:02:15 MST 2018

$ date; ./4ROLLdec ; date
Sat Mar 17 09:03:12 MST 2018
outputting file to /tmp/output with size 1490615928
had 270563811 junk nucleotides
Sat Mar 17 09:04:40 MST 2018

bytes
1422974976   - gen2.new (gentoo_live.iso decoded)
5962463715   - gentoo.dna
1422974976   - gentoo_live.iso

 encoded   /   decoded
5962463715 / 1422974976
4.190139542552293 x original size

junk      /  total size
270563811 / 5962463715
0.04537785451328319  (4.5% junk)

===========
MLK STATS
===========

bytes
168539   -  MLK_excerpt_VBR_45-85.mp3
705890   -  new (MLK mp3 encoded)

outputting file to /tmp/output with size 176472
had 31734 junk nucleotides

encoded / decoded
705890  / 168539
4.188288764024944 x original size
70589

junk  / total size
31734 / 705890.
0.044956012976526086 (4.5% junk)

===========
PROS
===========

-fast
-no libraries or fancy gadgets
-can avoid defined length of repeating homopolymer nucleotides

===========
CONS
===========

-not fully tested
-uses 'garbage' value to compensate for homopolymer repeats (possible to use these for parity bit for data integrity?)
-single nucleotide error means data could not be reliable (possibly 'statically roll' table instead of 'rolling' based on nucleotides? using a data value outside the DNA could also allow for multi-threading)
-proof of concept quality (e.g. code has little error checking, will crash if you don't have enough memory to encode the file you're using)
-may have a size error in decoding (see code. bs_offset not checked for size properly)