99 comments

[ 5.1 ms ] story [ 152 ms ] thread
What kind of things can be done with this? Are there tools that consume genomes and tell you something about them? Is biology advanced enough that it would be possible to create such tools?

I see they have several different strains. I could imagine trying to find similarities or organize them into a tree, if that's not already done.

In fact that is exactly what is done. What you are describing is genome annotation and homology based phylogenetic analysis.
Super awesome. As a point of curiosity, when did this kind of complete sequencing capability get to the point where it could be turned around in under two months?
It's been a while. You can do real time sequencing now. E.g., here is a sequencing demo happening in a coffee shop in Africa. https://twitter.com/virology_chitra/status/12200656783272550... The device connected to the laptop via USB is the sequencer.
The company that makes that is Oxford Nanopore https://nanoporetech.com/products
This is amazing. I'm having to restrain the part of myself that loves buying gadgets. These not only look cool, from watching their videos and reading about them, they have an awesome function.

Suppose I had some of this equipment and a sick family member. Would it be possible for me to somehow isolate a virus or bacteria from the sick person's saliva or blood and then sequence the DNA/RNA and then use a catalog of genetic information to identify a likely match?

This already has been available for some time. I found it through a link on the wikipedia page. I had not see the tree before. Because it is close to coronoviruses found in bats, one could conclude that bats are the animal reservoir for this type of virus (as was the case with SARS as well). I presume that China will forbid the trade and consumption of bats.
I also noticed the article several days ago. Just wondering if there is any basic tutorial in python to analyze this type of data.
I guess there is not much to analyze. Both the RNA sequence is given and their translation into amino-acids. The capital letters give the amino-acids. https://en.wikipedia.org/wiki/Amino_acid The lower case letter describe the RNA sequence. Sequences of three RNA base pairs are translated to one amino-acid. A table giving the standard decoding, can be found at: https://en.wikipedia.org/wiki/Genetic_code#RNA_codon_table

The next step is taking the amino-acids sequence and determine the protein structure. amino-acids chains fold together in a complex structure, which determine their function. Determining how this folding happens is a very complex problem. If you want to get some more idea about it, I suggest you have a look at https://fold.it/

Do you think research on 2019-nCoV would benefit from having available a platform for distributed computing, e.g. a dedicated Folding@Home?
I am afraid not. If there would be any, I guess it would already have been organized. No, knowing the structure of the proteins that make up the virus, might give some understand about how it works. If that gives some information about prevention (developping a vaccine) or treatment (developing anti-viral drugs), I wonder. Knowing the genetic code, can be helpful with developing vaccine. I understand that one approach is to incorporate part of the DNA in a known 'disabled' virus and test if it has some protective effect in some animals, and only after positive results test it on humans.
If I recall, the widely circulated video of the TV host eating bat soup was actually in Palau.

So even if China did ban it, would other countries have to as well for it to be effective?

Different species of bats have different degrees of similarity between human cell receptors. The edible bat soup may be a very different species that can’t transmit disease to humans.
You don't need to actually consume bats to have the viruses transmitted across species. Bat droppings could contaminate/infect some intermediary animal.
Could you get sick from the air if you were in a cave full of bats with the virus?
From what I read, the viruses live in the intestine in bats. So no, only through consumption of bats themselves or contact with their excretions.
If their excretions create an aerosol then you could breath it in.
If I were to guess I'd say it is from collecting guano to use as fertiliser. This practice is quite common and used to be so common that there is an expression in Norwegian for something selling very well, "Det selger som hakka møkk" [1], which means "It is selling like chopped shit".

[1] https://www.sprakradet.no/svardatabase/sporsmal-og-svar/som-... (Norwegian)

I hate saying this but I didn't expect some default-theme WordPress set up when I clicked on a link from nih.gov –with the comments section, post rating, and pingbacks enabled.
Well, the Higgs boson was presented ok Comic Sans... they’re scientists, not webdesigner agonizing on artisanal layouts ;)
I bet this was done on purpose with tongue in cheek.
I’ve been in particle physics and I wouldn’t be so sure
(comment deleted)
Think of it this way:

They launch with near absolute MVP. Design is not the priority here; the information is. When it comes to dealing with a pandemic, every minute counts.

The default theme is a sign that someone who knows software and project management is working on it.

And to add to that, domain experts who will work on this information are probably used to obtaining new knowledge reading email newsletters. They'll be fine.
Except this blog is up since 2013, though. :) I think .gov sites have some style guide.
.gov sites usually have some style guide, but it often doesn't extend as far as fifth-level subdomains delegated to small teams.

Also, the default theme is very clean and readable. It's hard to do better than that, so I'm glad they left it as it is.

I found out a few days ago when an, interesting, related question was posted on bioinformatics.stackexchange.com : Why does the Wuhan coronavirus genome end in aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa (33 a's)? [0]

[0]: https://bioinformatics.stackexchange.com/questions/11227/why...

(comment deleted)
Reminds me of the skit about the Monty Python creature of the name "The Black Beast of Aaaaaaargh."
(comment deleted)
(This is baseless speculation by someone who has only taken Genetics 101)

Human cell nuclei produce mRNA from DNA, which then makes it's way to the ribosome to be expressed as a chain of amino acids, or a protien. On it's way to the ribosome, it needs protection from the cell's RNA cleanup enzymes, so it leaves the nucleus with a 5' cap and a poly-A tail.

Therefore, it makes sense that a virus that wishes to hijack human ribosomes would end in a poly-A tail (a bunch of As) in order to not get destroyed by human cells.

(comment deleted)
The virus propagation was said to have started in a market that sells, among other products, snakes.

2019-nCoV can go from bats to snakes to humans. Eating snakes or bats seems like a bad idea.

Previous viruses came from pigs and chicken and cows. Unless you also advise to stop eating those this just seems a bit of a stretch.
Sir, do you have a minute to talk about our lord and saviour, veganism?

I’m only half kidding. It makes sense to go vegan/vegetarian during an outbreak of diseases spread by consuming meat.

There is an enourmous effort put into ensuring the safety and quality of the meat products you have mentioned.
Seems like someone's never left a first-world country before...
Hah.

The US poultry industry is so bad at looking after their animals they have to wash carcasses in chlorine to even _approach_ safe levels of pathogens like salmonella and campylobacter.

The USDA is implementing a plan where the companies inspect themselves with little oversight from anyone. Maybe in Europe it's better.
I'd assume the effort is similar between bats and snakes being sold in an open air market in a Chinese village and chicken, pigs and cows being sold in an open air market in a Chinese village.
There's an interesting thought.
Honestly, I don’t think it was transmitted by eating these wild animals. This might just be a bunch of nonsense.

When you cook something, it tends to kill the virus and germs. Or at least makes it safer to consume. But, you can still get sick from the fluidic contact, like getting salmonella from handling raw chicken.

Let’s take a look at some scientific evidence.

Ebola was theorized to come from bat feces, that might have fallen on a toy, and a girl put it into her mouth. She didn’t eat the bat. (This is only a theory, as they don’t know the true origin.)

The MERS virus from the Middle East, was thought to come from sick camel. I haven’t heard of the locals there eating camel, so it must have transmitted via fluidic contact. (Again, this is only a theory, as they don’t know the true origin.)

SARS was theorized to come from bats as the carrier, but the vector was a civet. Maybe the infected fluids of the civet jumped to humans. (And again, this is only a theory, as they really don’t know the true origin.)

This 2019-nCoV is again thought to come from bats, but the vector is currently unknown.

So with all that, I think perhaps, a bat was a carrier, and got some other animal sick, which became the vector. And when they slaughtered the sick animal, the infected fluid must’ve infected a worker, who got sick. And this worker might’ve infected an old person, who has a weaker immune system, and got sick, who then transmitted it to others, and got them sick, and now we have a full-blown pandemic.

So, the true cause here, might not be that people ate the wild animal, but that the humans came into contact with infected fluids from wild animals. And the presence of bats and wild animals, at this Wuhan market, just increased the probability that an outbreak like this could happen.

As a thought experiment, in Texas, there are a bunch of bats in Austin. What if one of these bats are also a carrier of some disease, and they fly over a nearby chicken farm, and poops on the chicken, which gets the chickens sick. The chickens are now the vector. And now, these sick chickens are handled by humans, and the above scenario plays out in Texas. And then, a pandemic spreads out from Texas. Would you think that Americans are disgusting for consuming chickens here, if this theoretical scenario were to happen?

The likelihood of a mammal-to-bird-to-human infection chain ought to be lower than mammal-to-human. Nevertheless all meat handling in the industry is at least theoretically subject to high standards of hygiene. So if at all, handling the meat of wild mammals without proper hygiene is the problem here. And yes, that is somehow disgusting.
Any way anyone else can 'contribute' here to reaching a cure? I'm thinking - somehow bootstrapping a node to join a distributed blockchain network of protein folding simulations ...
I don’t think a blockchain would really help with task specific protein folding simulation stuff?
There's projects that use blockchain and cryptocurrency as a form of incentives and payment for resources that people are putting forward for computing power.
Money that would be better spend on super computers i think
I think any contributions are be welcomed.
You don’t need either of those for incentives and payment for resources that people put forward.
What other alternatives are there?
Cryptocurrency isn’t exactly the only way to transfer money.
Nice dodge from answering the question.
Well, you could just use regular old bank transfers. I just assumed that I don’t need to tell you this.

And even if you were using cryptocurrency for payouts, that hardly requires any blockchain stuff on the side consuming the computing power.

Join a company working in the space. There is a need at all of them for good UI, automation, and database engineers.
Happen to know of any off the top of your head?
...you mistyped an extra word in that sentence: "blockchain".
29,903 base pairs (4 bit values), so that’s 119,612 bits, or about 15 KB of information total. A devious little self-replicator in 15 KB which runs on our bio substrate!
I wonder how many of those base pairs are devoted to evading the host's immune system, as opposed to the primary function of being a fork bomb.
> opposed to the primary function of being a fork bomb

/me realizes that every organism is a fork bomb

Probably not that many, it should be sufficient that the immune system doesn't have anything to specifically deal with it, so it only has to evade the complement system (which for a virus isn't terribly hard). It also helps that the strategy is simply "replicate and just swarm the defense".
There are 4 base types (A,T,C,G) so it's only 2-bit values. There is also some redundancy coding in the amino acid sequence that follow the DNA. So if you compute the information of the codons sequence it's even lower.

It's quite compact. What would be interesting to know, is the percentage of programs still running when a few bits are modified, to know how sharp is the search space.

Exploring living programs space, you probably can use some advance neural network techniques, like GPT-2 to explore it more efficiently, directly in its compressed representation to provide an high level interface with slider controlling the high-level properties.

Gene overlap is also possible and exists in viruses, so the true amount of data is actually usually a bit larger than the the sequence's length because of this "compression".
"Scientists find Arbitrary Code Execution vulnerability in Human Cells, vendor unresponsive to security issue"
29903 base pairs is 59806 bits or 7.3 KB.

Hepatitis B is 800 bytes for comparison.

Kudos to the scientists who developed this so fast. I rember it took four YEARS to identify the HIV virus and another two years for a test.
It's really amazing how much more powerful biological tools have become since then!
Does anyone know of an easy way to download the nucleotide sequences for the entire set of pneumonia / coronavirus viruses? I've looked at the FTP mirror but can't connect it with the nucleotide locus I find on the site itself.

I work in language modeling and want to see about using those unsupervised / self-supervised methods for genome annotation / phylogenetic tree construction.

Even if it's more of a curiousity than a useful tool I have experience with small datasets, most recently focused on character level language modeling on ~90MB of training data, so if I can get (90MB / (29 kilobases * 2 bits per base) =) approximately 12,000 related samples I should be able to at least make a dataset out of it.

Using Biopython. Note that the search query that I am using currently returns 70605 results, so you might want to tweak it fit your needs.

    from Bio import Entrez
    import time
    from urllib.error import HTTPError
    
    DB = 'nucleotide'
    QUERY = '("pneumoviridae"[Organism] OR "Coronaviridae"[Organism])'
    
    Entrez.email = 'your.email@provider.com'
    handle = Entrez.esearch(db=DB, term=QUERY, rettype='fasta')
    record = Entrez.read(handle)
    
    handle = Entrez.esearch(db=DB, term=QUERY, retmax=count, rettype='fasta')
    record = Entrez.read(handle)

    id_list = record['IdList']
    count = len(id_list)
    post_xml = Entrez.epost(DB, id=",".join(id_list))
    search_results = Entrez.read(post_xml)
    
    webenv = search_results['WebEnv']
    query_key = search_results['QueryKey']

    batch_size = 200
    with open('viruses.fasta', 'w') as out_handle:
        for start in range(0, count, batch_size):
            end = min(count, start+batch_size)
            print(f"Going to download record {start+1} to {end}")
            attempt = 0
            success = False
            while attempt < 3 and not success:
                attempt += 1
                try:
                    fetch_handle = Entrez.efetch(db=DB, rettype='fasta',
                                                 retstart=start, retmax=batch_size,
                                                 webenv=webenv, query_key=query_key)
                    success = True
                except HTTPError as err:
                    if 500 <= err.code <= 599:
                        print(f"Received error from server {err}")
                        print("Attempt {attempt} of 3")
                        time.sleep(15)
                    else:
                        raise
            data = fetch_handle.read()
            fetch_handle.close()
            out_handle.write(data)
Champion! Thank you =]
There is a small error in the code. The variable `count` should be defined on line 11 like:

    count = int(record["Count"])
en the appearance on line 15 should be removed.
I wonder how realistic it is to expect a vaccine in 3 months, given that we don't have a vaccine for MERS 8 years later (and not really one for SARS either)? Is it just a question of allocating resources and shifting priorities?
This is an interesting question, and I can confidently say the technology exists to have some quantity of human vaccine within 3 months. Will it be enough to have an impact? Probably not. We're talking about needing millions of doses. But if the companies in the space can succeed at at least showing the turn-around-time for some quantity of working vaccine is possible, then it becomes an engineering problem and not a scientific problem. Then you just need to marshal resources and you could be at a place where bulk vaccine could be produced in that short window of time. This 2019-nCoV marks an inflection point for humanity IMHO where sequencing costs, data transmissions, and vaccine technology are all good enough to actually accomplish a real-time vaccine. The science is here, now it's just turning a bigger crank.
I read they have a few SARS vaccines, but non that have gone through human trials yet.

Side not: I'm always concerned when we're put in a situation like this where we're pressured to put out a vaccine quickly because it opens the door for a less than optimal solution to become a mainstay. For example since it's an emergency you may put forward a solution, have it go through human trials, find some side effect, but given the current circumstance say that it's "good enough" and give it the green light of approval. Now I understand that in such an urgent circumstance that this may be justified, but after the urgency is over it's my understanding that this solution that may or may not be approved under regular circumstances will remain approved.

Why should a given vaccine have more side effects than existing vaccines? It’s just broken up pieces of the virus. If produced properly and rendered inert Why should it trigger major side effects?
More interesting to me is the process the medical industry has to go through to deal with the testing kits. Apparently they are in short supply. Lots of good nuggets in this thread. Yes, it's Twitter, but this can be cross-checked.

https://twitter.com/luchenhist/status/1220497118755987456

* Testing is a long process * Kit distribution is supply chain constrained due to holiday * Only severe cases are being tested * Hospitals are overwhelmed * People not getting tested are being turned back * Government offering compensation for testing but only if it's actually Coronavirus

The author has other interesting Tweets as well. Might be worth a follow if this is of interest.

(comment deleted)
(comment deleted)
Looks like the Chinese military had the same Bat SARS-like Coronavirus in their possession 2 years ago.

Here is the NIH page showing Identical Protein Matches of the envelope protein from Wuhan 2019-nCoV:

https://www.ncbi.nlm.nih.gov/ipg/QHO62113.1

Notice there are two entries not from the recent Wuhan seafood market outbreak:

https://www.ncbi.nlm.nih.gov/protein/AVP78033.1

https://www.ncbi.nlm.nih.gov/protein/AVP78044.1

Both sequences were provided on 05-JAN-2018 by:

  Institute of Military Medicine Nanjing Command, 
  NO. 293 East Zhongshan Road, Nanjing, JangSu 210002, China