larlet-fr-david-cache/cache/2021/5f1c22e9a41d209ff84218b3d6faf676/index.md

title: Reverse Engineering Source Code of the Biontech Pfizer Vaccine: Part 2
url: https://berthub.eu/articles/posts/part-2-reverse-engineering-source-code-of-the-biontech-pfizer-vaccine/
hash_url: 5f1c22e9a41d209ff84218b3d6faf676

<p>All BNT162b2 vaccine data on this page is sourced from this <a href="https://mednet-communities.net/inn/db/media/docs/11889.doc" target="_blank">World Health
Organization
document</a>.</p>

<blockquote>
<p>This is a living page, shared already so people can get going! But
check back frequently for updates.</p>
</blockquote>

<p><em>Translation</em>:
<a href="https://renaudguerin.net/posts/partie-2-explorons-le-code-source-du-vaccin-biontech-pfizer/" target="_blank">Français</a>
/ <a href="https://msakai.github.io/bnt162b2/part-2-reverse-engineering-source-code-of-the-biontech-pfizer-vaccine.ja/" target="_blank">日本語</a></p>

<p>In short: the vaccine mRNA has been optimized by the manufacturer by
changing bits of RNA from (say) <code>UUU</code> to <code>UUC</code>, and people would like to
understand the logic behind these changes.  This challenge is quite close to what
cryptologists and reverse engineering people encounter regularly.  On this
page, you’ll find all the details you need to get cracking to reverse
engineer just HOW the vaccine has been optimized.</p>

<p>I thought this would just be a fun puzzle, but I have just been informed that
figuring out the optimization procedure &amp; documenting it is tremendously
important for researchers around the world, as this would help them design
code for proteins and vaccines.</p>

<p>So, if you want to help vaccine research, do read on!</p>

<h2 id="the-leader-board">The leader board</h2>

<p>Here are the current best entrants to the optimization algorithm (average of 20 runs):</p>



<h2 id="biontech">BioNTech</h2>

<p>We should all be very grateful that BioNTech has shared this data with us.
And of course we should also be grateful to the many many researchers and
lab workers that worked for decades to bring the state of the art to the
point that such a vaccine could be developed.  It is marvelous.</p>

<p>Because it is so marvelous, I want to understand everything about the
vaccine. I wrote a page <a href="https://berthub.eu/articles/posts/reverse-engineering-source-code-of-the-biontech-pfizer-vaccine/" target="_blank">Reverse Engineering the source code of the BioNTech/Pfizer SARS-CoV-2
Vaccine</a>
that describes in some detail what is in the mRNA of the vaccine. It helps
to read this page before continuing, I promise you it will be interesting.</p>

<p>The post left open some questions however, and this is where it gets
fascinating.</p>

<h2 id="the-codon-optimization">The codon optimization</h2>

<p>The vaccine contains RNA code for a very <em>slightly</em> modified copy of the
SARS-CoV-2 S protein.</p>

<p>The RNA code of the vaccine itself however is <em>highly</em> modified from the viral original!
This has been done by the manufacturer, based on their understanding of
nature.</p>

<p>And from what we understand, these modifications make the vaccine <strong>much
much more</strong> effective.  It would be a lot of fun to understand these
modifications.  It might for example explain why the Moderna vaccine needs
100 micrograms and the BioNTech vaccine only 30 micrograms.</p>

<p>Here is the beginning of the S protein in both the virus and the BNT162b2
vaccine RNA code.  Exclamation marks denote differences.</p>

<pre><code>Virus:   AUG UUU GUU UUU CUU GUU UUA UUG CCA CUA GUC UCU AGU CAG UGU GUU
Vaccine: AUG UUC GUG UUC CUG GUG CUG CUG CCU CUG GUG UCC AGC CAG UGU GUG
               !   !   !   !   ! ! ! !     !   !   !   !   !            
</code></pre>

<p>RNA is a string (literally) of RNA characters, <code>A</code>, <code>C</code>, <code>G</code> and <code>U</code>. There is no
physical framing on there, but it makes sense to analyse it in groups of
three.</p>

<p>Each group (called a codon) maps to an amino acid (denoted by a capital
letter).  A string of amino acids is a protein.  Here is what that looks
like:</p>

<pre><code>Virus:   AUG UUU GUU UUU CUU GUU UUA UUG CCA CUA GUC UCU AGU CAG UGU GUU
          M   F   V   F   L   V   L   L   P   L   V   S   S   Q   C   V
Vaccine: AUG UUC GUG UUC CUG GUG CUG CUG CCU CUG GUG UCC AGC CAG UGU GUG
               !   !   !   !   ! ! ! !     !   !   !   !   !            
</code></pre>

<p>Here we can see that while the codons are different, the amino acid version
is the same. There are 4*4*4 codons but only 20 amino acids. This means you
can typically change every codon into one of two others, and still code for
the same amino acid.</p>

<p>So in the second codon, <code>UUU</code> was changed to <code>UUC</code>. This is a net addition
of one ‘C’ to the vaccine. The third codon changed from <code>GUU</code> to <code>GUG</code>, which is
a net addition of one <code>G</code>.</p>

<p><strong>It is known that a higher fraction of <code>G</code> and <code>C</code> characters improves the
efficiency of an mRNA vaccine</strong>.</p>

<p>Now, if that was all there was to it, this could be the end of this page.
“The algorithm is change codons so we get more G and C in there”. But then
we meet the 9th codon which changes <code>CCA</code> to <code>CCU</code>.</p>

<p>Throughout the ~4000 characters of the vaccine, this happens many times.</p>

<h2 id="our-challenge">Our challenge</h2>

<p>The goal is: find an algorithm that modifies the ‘wild type’ RNA code into
the BNT162b2 one. Because everyone would like to understand how to turn
viral RNA into an effective vaccine. The algorithm does not need to
reproduce the <em>exact</em> RNA code of course, but it would be super nice if it
came up with something very similar, while also being brief.</p>

<p>To help you, I have provided the data in a number of forms, as described on
<a href="https://github.com/berthubert/bnt162b2" target="_blank">the GitHub page</a>.</p>

<blockquote>
<p>Note that in these files the <code>U</code> mentioned above appears as a <code>T</code>. <code>U</code> and
<code>T</code> are the RNA and DNA manifestations of the same information.</p>
</blockquote>

<p>The easiest place to start might be the
‘<a href="https://github.com/berthubert/bnt162b2/blob/master/side-by-side.csv" target="_blank">side-by-side.csv</a>‘
file. This lists the original and modified version of each codon, side by
side:</p>

<pre><code>abspos,codonOrig,codonVaccine
0,ATG,ATG
3,TTT,TTC
6,GTT,GTG
...
3813,TAC,TAC
3816,ACA,ACA
3819,TAA,TGA
</code></pre>

<p>There is also an equivalency table that shows wich codons can be
interchanged without changing the amino acid output. Please find this in
<a href="https://github.com/berthubert/bnt162b2/blob/master/codon-table-grouped.csv" target="_blank">codon-table-grouped.csv</a>.
There is also a visual version
<a href="https://en.wikipedia.org/wiki/DNA_and_RNA_codon_tables#Standard_DNA_codon_table" target="_blank">here</a>.</p>

<h2 id="a-sample-algorithm">A sample algorithm</h2>

<p>On the <a href="https://github.com/berthubert/bnt162b2" target="_blank">GitHub repository</a> you can
find
<a href="https://github.com/berthubert/bnt162b2/blob/master/3rd-gc.go" target="_blank">3rd-gc.go</a>
(and
<a href="https://github.com/berthubert/bnt162b2/blob/master/3rd-gc.py" target="_blank">3rd-gc.py</a>).</p>

<p>These implement a simple strategy that works like this:</p>

<ul>
<li>If a virus codon already ended on G or C, copy it to the vaccine mRNA</li>
<li>If not, replace last nucleotide in codon by a G, see if the amino acid
still matches, if so, copy to the vaccine mRNA</li>
<li>Try the same with a C</li>
<li>Otherwise copy as is</li>
</ul>

<p>Or in <code>golang</code>:</p>

<pre><code>// base case, don't do anything
our = vir

// don't do anything if codon ends on G or C already
if(vir[2] == 'G' || vir[2] =='C') {
	fmt.Printf("Codon ended on G or C already, not doing anything.")
} else {
	prop = vir[:2]+"G"
	fmt.Printf("Attempting G substitution, new candidate '%s'. ", prop)
	if(c2s[vir] == c2s[prop]) {
		fmt.Printf("Amino acid still the same, done!")
		our = prop
	} else {
		fmt.Printf("Oops, amino acid changed. Trying C, new candidate '%s'. ", prop)
		prop = vir[:2]+"C"
		if(c2s[vir] == c2s[prop]) {
			fmt.Printf("Amino acid still the same, done!")
			our=prop
		} 
		
	}

}
</code></pre>

<p>This achieves a rather poor 53.1% match with the BioNTech RNA vaccine, but
it is a start.</p>

<p>When you design your algorithm, be sure to only base your choices on the
virus RNA. Do not peek into the BioNTech RNA!</p>

<p>If you have achieved a score beyond 53.1% please email a link to your code
to bert@hubertnet.nl (or <a href="https://twitter.com/PowerDNS_Bert" target="_blank">@PowerDNS_Bert</a>
and I’ll put it on the leader board at the top of this page!</p>

<h2 id="things-that-will-help">Things that will help</h2>

<p>As with every form of reverse engineering or cryptanalysis, it helps to
understand what we are looking at.</p>

<h2 id="gc-ratio">GC ratio</h2>

<p>We know that one goal of the ‘codon optimization’ is to get more <code>C</code>s and
<code>G</code>s into the vaccine version of the RNA. However, there is also a limit to
that. In DNA, which is also used to manufacture the vaccine, <code>G</code> and <code>C</code>
bind together strongly, to the point that if you put too many of these
‘nucleotides’ in there, the DNA will no longer be replicated efficiently.</p>

<p>So some modifications may actually happen to manage <em>down</em> the GC percentage of a
stretch of DNA if it was getting too high.</p>

<p>I <a href="https://twitter.com/PowerDNS_Bert/status/1344036143961169920" target="_blank">tweeted about this</a> earlier.</p>

<h2 id="codon-optimization">Codon optimization</h2>

<p>Some codons are rare in human DNA, or in certain cells. It may be that some
codons are replaced by other ones simply because they are more frequently
used by some cells.</p>

<p>I <a href="https://twitter.com/PowerDNS_Bert/status/1344400081802448897" target="_blank">tweeted about this</a>
earlier.</p>

<h2 id="rna-folding">RNA folding</h2>

<p>We’ve been looking at codons up to here. The RNA itself however does not
know about codons, there are no markers that say where a codon begins and
ends. The first codon on a protein however is always ATG (or AUG in RNA).</p>

<p>RNA curls up into a shape. This shape might help evade the immune system or
it might improve translation into amino acids. This only depends on the
sequence of RNA nucleotides and not on specific codons.</p>

<p>You can submit RNA sequences to <a href="http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi" target="_blank">this server of the Institute for
Theoretical Chemistry at the University of
Vienna</a> and it
will fold RNA for you.  This is a very advanced server that does meticulous
calculations.</p>

<p>This <a href="https://en.wikipedia.org/wiki/Nucleic_acid_structure_prediction" target="_blank">Wikipedia
page</a>
describes how this works.</p>

<p>It may be that some optimizations improve folding.</p>

<p>I am also told that this paper by Moderna (another mRNA vaccine
manufacturer) may be relevant:
<a href="https://www.pnas.org/content/116/48/24075" target="_blank">mRNA structure regulates protein expression through changes in functional
half-life</a>.</p>