thethoughtemporium/whose-gene-is-it-anyway

Whose-gene-is-it-anyway

Welcome to the "Whose Gene is it Anyway" Repo

This is the collection of all the plasmids designed during the "whose gene is it anyway" live streams.

These plasmids are released as creative commons attribution-share alike. You may download them, modify them, even have them synthesized if you so choose or use them for commercial activity. But you must provide attribution to The Thought Emporium and Justin Atkin, and any derivatives you make must be released under the same license. See the link below for details. https://creativecommons.org/licenses/by-sa/4.0/

Link to channel: https://youtube.com/c/thethoughtemporium

Last stream: https://youtu.be/20vwKHfbVvY

All streams: https://www.youtube.com/playlist?list=PLZLsjPxmF1BEI5CReoklVP4u84kMkjIZp

whose-gene-is-it-anyway's Issues

Potential bug in the Definitely not zombie serum v3 aav IgGA code?

I was wondering if the sequence posted as the V3 is the up to date sequence since it has been a while since upload? I was looking through the sequence and there are 2 NdeI cut sites, (the one you added and one on the AAV backbone) is this something that has been fixed previously?

• I am assuming you didn't change the NdeI cut site between the rVAR2 and the IgGA when you switched from a lentiviral to an AAV system

If not fixed it may be a good switch since, if you wanted to swap the human IgGA FC to another FC. For example, to a mouse IgG for testing in mice (I am relatively sure that human FC's won't work in mice but I could be mistaken) then it will be much easier to just synthesis the mouse IgG and clone it after removing the human IgGA. As the plasmid stands now you will have to either remake your insert with a mouse IgG or cut your plasmid with BspEI and have your insert repair the rVAR2 C-terminal sequence with a new IgG FC.

If I am correct, hopefully, you haven't ordered the synthesis of this insert yet but either way, I hope this helps!

PS. I am currently doing my MSc in Plant Biochemistry and your videos on benching have helped me save so much time. So thanks for posting these! and keep up the good work

Add a LICENSE.md file

TL;DR

Consider adding a LICENSE.md file to the root of this repository.

By using this name GitHub will automatically be able to detect which one it is and will display it as a badge above the root rile listing (next to the contributors count).

About the mention in the README.md

I realize that the README.md file of this project already references the CC BY-SA 4.0.

This issue addresses 1) the need for a LICENSE.md file, and 2) the suggestion to choose a different license.

Furthermore, if someone creates a derivative work of your project, they may license it under the terms of the GNU GPL 3, but you won't be able to include their improvements in the original project, because it would violate the GPL (at least to my understanding, I'm not a lawyer) as stated below:

"CC BY-SA 4.0 is one-way compatible with the GNU GPL version 3: this means you may license your modified versions of CC BY-SA 4.0 materials under GNU GPL version 3, but you may not relicense GPL 3 licensed works under CC BY-SA 4.0.
- GNU licenses list, section "ccbysa"

Choosing the right license

I would highly recommend to choose something like the GPL 3 or something like it.
If you're ok with corporations using this stuff without disclosing their changes, then go for the MIT license instead.

I'll not go into any more detail, because there already is a very good resource created by GitHub themselves called "Choose a license":

https://choosealicense.com/

My personal favourite is the AGPL 3 license (of which an official markdown version can be found on the GNU website: https://www.gnu.org/licenses/agpl-3.0.md).
It's a license taylored to sharing improvements and empowering user freedom (which I think is important, especially for a project about gene editing).

---

Thanks for reading my wall of text;
I'd love to hear your opinion on this.

Kozac Sequences for Spicy Tomatoes

So I've noticed that when you open the pcambia2300_spicy_tomato file in benchling, there is no kozac sequence for the plasmid. From what I understand, it's something that must be present in lentiviruses and retroviruses like agrobacterium (what the pcambia2300 plasmid backbone is made for) in order for any sort of transcription or gene expression to occur. This may be a problem that stops any new genes from being expressed. I don't really have the money to drop on an experiment like this right now but if anyone has tried this one before could you please share your results and whether or not you ran into any issues with this? Thank you in advance!

Malaria cancer therapy - things that makes this hard?

Hey,

Firstly, I very appreciate your work and that you inspire even more people to learn and pursue hard problems.

I watched whole stream and it seemed just too easy. I understand that I don't have enough knowledge here, but I would really like to know what would make this hard/impossible. I recently got diagnosed with cancer and there are some things that just bug me after watching you stream.

• Is it the immunity that 95% of time just kills of injected virus? And then scientists have to iterate and do slight randomish changes until it works?
• Is there some hard constraints that plasmid must follow for it to work and it is hard computing to get it to work?
• Are trails so long and expensive that most of ideas don't get anywhere?
• Are humans so different from each other that there is just too much risk?
• Are you the first one who come up with idea to use malaria + hiv?
• Why wouldn't multi billion dollar companies just throw money on this and iterate over all possible plasmids that could achieve this? (patents?) Or they have but everything failed. (But then why)

p.s. It seemed funny to me how creating plasmid looked very similar to hacking with ROP (https://en.wikipedia.org/wiki/Return-oriented_programming) to me.

Codon Shuffle code

Why don't you use BioPython? With it, you don't need to deal with external codons tables and checks and it has proper instruments for dealing with sequences. You can also use specific codon tables.

A toy code example:

import random
import numpy as np
from Bio.Data import CodonTable
from Bio.Seq import Seq
random.seed(2128506)

def new_random_DNA(codons, stop_codons, protein_length): 
	'''
	Generates random DNA sequence with length 3*protein_length from codons list and without
	items from stop_codons list
	'''

	random_DNA = []
	checkLengthFlag = False
	while not  checkLengthFlag:
		random_codon = ''.join(random.choices(codons, k=3))
		if random_codon not in stop_codons:
		random_DNA.append(random_codon)
		if len(random_DNA) == 3 * protein_length:
		checkLengthFlag = True
	random_DNA =''.join(random_DNA)
	return random_DNA

def random_DNA_shuffle(seq_DNA, aa_coding, times=1):
	'''
	Shuffles DNA sequence while keeping protein sequence intact
	'''
	seq_DNA = Seq(''.join(seq_DNA))  
	protein = seq_DNA.translate()
	aa_list = list(protein)
	new_DNA_list = []
	for _ in range(times):
		new_DNA = ''
		for aa in aa_list:
			new_codon = random.choice(aa_coding[aa])
			new_DNA += new_codon
		new_DNA_list.append(new_DNA)
	return new_DNA_list


aa_coding = dict()
for key in _temp:
	aa_coding.setdefault(_temp[key], []).append(key)

table = CodonTable.unambiguous_dna_by_name["Standard"] 
stop_codons = table.stop_codons
codons = table.nucleotide_alphabet
protein_length = 10

aa_coding = dict()
for key in _temp:
	aa_coding.setdefault(_temp[key], []).append(key)

seq_DNA = new_random_DNA(codons, stop_codons, protein_length)
new = random_DNA_shuffle(seq_DNA, aa_coding, times=1)

print(random_DNA) # preview 
for seq in new:
	print(seq)
print(Seq(random_DNA).translate())
for seq in new:
	print(Seq(seq).translate()

Forgive me - but please read - i think you'll like this

I couldn't find a way to message you directly - so i am placing this here. I ran into this white paper recently: https://corescholar.libraries.wright.edu/cgi/viewcontent.cgi?article=2831&context=etd_all

I have been going through the paces of attempting to replicate this project - as the potential is huge. Do to the similarity of the polymer-chains to Drag-line silk - and the given the assembly process of the polymer is similar - I was thinking this would make an excellent Whose-gene-is-it-anyway live stream :). Please, check out the paper and let me know what you think. Also - i know this isn't an issue - so i apologize.

Extending scripts: what's needed / useful?

I'm a programmer who would like to contribute, but I don't know how best to help. If someone can make a list of programs that they would like to have written, we can make it happen.

A few ideas for scripts that could be useful:

1. A script to check the "printability" of a particular gene sequence -- to see how much it repeats, and identify problem sections that may need to be scrambled

2. A script to visualize the resulting sequence. At 14 minutes in the video, you gave the SR-71 Blackbird example -- would it help to have a tool to let you visualize the original sequence alongside the new sequence to make sure it still looks "silky"?

3. A better interface for taking a sequence and shuffling / extending it? The existing command-line application is good and all, but it might be nice to wrap a web-interface around it? If someone who understands the application can napkin-sketch a user interface, I can make it happen -- I just need to know roughly what it should look like, and we can infer / define the behavior.

Are any of these scripts useful? Are there other ones that would be better?

Sorry for my lack of domain-specific knowledge, but I really would love to help with this sort of thing, and building a powerful set of open-source tools to help you or others in this domain is something I would love to help with.

Thank you!

Questions about attribution

I am aware that the world of Licensing is one which most people don't want to deal with. It's purpose of it in Open Source is to provide certainty to people who want to use/change it.

Using CC as a license for source code is discouraged. This is due to software being significantly different from other media. Some day there might be a plasmoid license but until then a few open questions remain around attribution.

Your license of choice requires:

Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.

Could you give your opinion those scenarios about the deer milk plasmoid (this should cover most unclear scenarios):

• Someone takes the unchanged plasmoid and has it produced and redistributes vials of the DNA in water. Would you want people to put a attribution statement on the vial or pass along a paper with it?

• Someone takes the unchanged plasmoid and puts it in an organism and redistributes vials of the organism. Would you want people to put a attribution statement on the vial or pass along a paper with it?

• Someone takes yeast containing the unchanged plasmoid and runs it, they don't kill/filter of the yeast and give away the milk (starter) in containers. Would you want people to put a attribution statement on the bottles or pass along a paper with it?

• Someone takes yeast containing the unchanged plasmoid and runs it, they kill/filter of the yeast and give away the milk away in containers. Would you want people to put a attribution statement on the bottles or pass along a paper with it?

Upload v2 of Spider Silk plasmid

Hey there,
it looks like the plasmid in the repo is the old one with the wrong mineralization stuff. Would you mind sharing the v2 with the fixes?

Also, does anybody know about where the spider silk project is at currently? I didn't see any videos, and I haven't watched through the livestreams or sifted through social media etc yet.

Would a de Bruijn sequence be more use for codon shuffling?

I was wondering if a de Bruijn sequence would be more suitable than a random shuffle. This lets you put constraints on things like minimum periodicity or stretch things out to the maximum possible period within defined constraints.

That said, am I right in understanding that you want to avoid not only repetitive sequences but also sequences that would be repetitive according to some sort of a fuzzy match? If that is a concern then there's probably a way to achieve this as a de Bruijn sequence but it's not prescribed explicitly in that definition.

BTW, while writing this I stumbled across something that looks relevant but I don't understand. I know absolutely nothing about genomics. Maybe this is an old idea. 🤷

Cancer immunotherapy might cause allergies against placenta

I'm not a Biologist, just a curious laymen so take the following with a grain of salt.

If I understand the concept of the "malaria cancer therapy" project correctly it basically tries to make your immune system react to cancer cells with a normal immune reaction.
The marker you're using however is the same as expressed in human placenta cells.

Wont this "treatment" then basically make your immune system also permanently react against your placenta (if you have one).
So in effect people treated with this would develop an allergy against their own placenta, rendering them infertile?

If this reasoning holds this type of immunotherapy would be an incredibly stupid idea...

Say the most common repeats in codonscrombler analytics

"also building in some analytics would be great...and eventually some way of seeing how often repeats of 10 letters or more show up"
-thethoughtemporium on discord

Drop Downs

Add a dropdown or something so people can switch between codon tables for organisms in codonscrombler.

Silk protein length

As you said in the video, your protein is much smaller than real ones. Maybe it's not a big deal, but what do you think about the SpyCatcher-SpyTag approach? The head-on polymerization approach can result in cyclic proteins (something like this), but there is a modification which allows bypassing this. In Fierer et. al 2014 authors developed a three-component approach - two tags and one "SpyLigase". All the sequences are in supplementary materials. They are quite small:

SpyLigase
MSYYHHHHHHDYDGQSGDGKELAGATMELRDSSGKTISTWISDGQVKDFYLYPGKYTFVETAAPDGYEVATAITFTVNEQGQVTVNGKATKGGSGGSGGSGEDSATHI
SpyTag
AHIVMVDAYKPTK
KTag
ATHIKFSKRD

So, basically, my idea is - you add KTag and SpyTag on N and C ends of your silk construction, modify SpyLigase to be produced by another strain of yeast out of cells, as with silk. Then you combine super from SpyLigase strain and silk suspension, and theoretically, it will polymerize (maybe it should be done on an ultrasound bath or something). Or you can grow these two strains in one bioreactor and silk will polymerize as it shows out of a cell.

P.S.
If there is a paywall, you can use this, but it's not particularly legal. Or I can upload those papers by myself in here.

Licensing problems with your repo

These plasmids are released as creative commons fair use. You may download them, modify them, even have them synthesized if you so choose. The only requirement is that you give attribution to The Thought Emporium and whoever was helping to design the plasmid during the stream.

Seems to approximate a license. However there are several it has several problems:

• "creative commons fair use" is undefined since there is no license with that name. Fair use is a USA specific legal concept which does not exist in different jurisdictions.
• There are several creative commons licenses which one is used matters.
• Would you be allowed to synthesize them yourself?
• How would attribution look like? Including your name in the plasmid? For software it is trivial since the length of a license is minimal compared to the code/binary. However for DNA this is not the case.
• In software it is usually seen as the maintainers task to maintain the list of all people who have contributed to the project. "whoever was helping to design the plasmid during the stream" is woefully undefined. Would you include the one who pointed you the base plasmid? The other people who pointed you plasmids who did not work? The guy who made the GNU/Milk joke? Would screen name on Youtube suffice or would researching there real identity be necessary? What if the stream or the chat log gets lost? I think under those circumstances it is impossible to reuse your work, which is not what you intend. Also it think it might be required to watch all your videos since there is currently no mapping between videos and plasmids.
• I am not sure if plasmids designs when handed out for free for come with implied guarantees in some jurisdictions. Software licenses almost always include boilerplates stating that the software is provided as is and without any warranties. Maybe if you should choose a license which includes such a disclaimer.

I am not very familiar with open source plasmid licenses however it seems that the scene is not very developed so choosing a (slightly modified) software license might be the way to go.

I could help you pick a license if you feel not comfortable to do that on your own.

Zombie Serum LacZ Alpha missing stop codon?

Disclaimer: I haven't studied biology, and I'm completely new to this.

LacZ Alpha appears to be missing a stop codon. EDIT: Specifically at 4795, I would have expected an R followed by a stop codon. Instead, f1 ori starts immediately at 4795 without anything separating it from the end of LacZ Alpha.

The definitely not zombie serum contains LacZ Alpha, except something (AAV packaging surrounding code to produce your cancer-tagging protein and to protect itself in a human cell) has already been inserted near the start of it, breaking it, which I assume was deliberate.

But if you're planning on putting this into cells that lack LacZ Alpha, then removing everything from pacI to pacI using restriction enzymes and then using blue/white screening to make the cells that have had the code successfully removed turn blue, then the missing stop codon might be a problem. (Or it might not, I have no experience with this.)

Silk project lacking resources

Would it be possible to upload your Google Doc containing some of the DNA sequences you used for the randomization step? If you don't mind, I'd like to write a little genetic randomizer program for you since I'm sure you'll be working on other versions of the silk project to get different properties.

Don't have money to donate but I do have programming ability and free time.