Despite all the sadness and disappointment regarding his misdeeds I can't resist saying the guy is talented in conveying interesting content in such a captivating way that he keeps me watching
Dude, THANK YOU. This is a great explanation! But also, the AlphaFold code available says it can't be used to predict structures outside of the CASP13 dataset. Does this mean you actually can't input your own sequence of interest, or just that accuracy cannot be ensured with other sequences?
Can you use this code to predict new protein structures? Where do you put the file with the sequence you want to predicted? Can you visualize using PyMOL?
(alphafold_venv) (base) jisna@ccc-CELSIUS-W570power:~$ pip install -r alphafold_casp13/requirements.txt Could not open requirements file: [Errno 2] No such file or directory: 'alphafold_casp13/requirements.txt' I am getting this error..
Watching this as a 50 year old GenXer...confirms that I totally dont give a fuck that young people didn't learn cursive. Ya'll know code. And code jargon. Communication tools orders of magnitude more powerful.
Siraj, we cannot use our own sequence to predict the structures using AlphaFold, as per the GitHub website data and the alphafold website. So, how you did it?
what if you accepted you don't know machine learning and AI you talk so much about. Go on to take a break and then coming back after a year or so with strong knowledge of concepts. just a recommend though i know u r a TH-camr and not a AI teacher in any way.
This is a huge breakthrough in biology. But saying that it will replace biochemists is like saying that a stat. analysis software will replace mathematicians. AI will play a huge role in future scientific research but that doesn’t mean that it will make biochemists useless.
They've not solved it, at CASP they correctly predicted 25 out of 43, and also they don't really know what's going on. They are just using similar structures to predict.
Sorry, but the original Alphafold implementation is a dead-end. Some others were made: 1- ProSPr (github.com/dellacortelab/prospr) 2- Alphafold_pytorch (github.com/Urinx/alphafold_pytorch) A paper on the first is in bioRxiv since November. I would rather focus on this one.
Hey Siraj, can you make a video on the Bayesian neural network? it's a very underrated topic in the field of machine learning but I feel like the intersection of the graphical model, neural network and Bayesian statistics will the next breakthrough
Hii Siraj, recently i am checking your video to understand about #auto image captioning and i am not able to found any video on this topic so if you are able to make video in this topic that will be great. BY the way loves video
Can que simulate simple cells? If yes, can we simulate simple multicelular organisms like a worm? If yes can we simulate complex multicellular organism like a fish and it's brain? Or a human brain?????
@@SirajRaval we are all made of cells right? So when it happens we would be able to simulate a entire human being and discover the secrets of the brain and Biological inteligence???
Dev do futuro , seems like it should be that simple, right? I learned how to cook an egg, now I know how to cook all dishes that contain eggs, I learned how to hold, tune, and pluck a guitar, hold on to your face Eddie Van Halen, I’m gonna out-shred you tomorrow. Weirdly, we can’t yet jump from fundamental principles to mastery that quickly, but soon Alpha-universe will. If a step takes a month, and it takes 50 exponential steps to get to 100 trillion, we should be on track to get to human intelligence this decade and all intelligence in 25 years, but not tomorrow or even in the first quarter of this amazing year.
Knock knock Whose there? I am the Protein folding problem, cause it's still not solved. Mind the difference between structure prediction and structure determination.
Siraj, there's a Git Issue that may help categorize which of your repositories can be fixed automatically and which that will require human intervention... github.com/git-utilities/fix_logs/issues/1#issuecomment-579574567 ... the default `config.json` has also been updated with those repositories that are easiest to fix via script. Let me know if questions on resolving merge conflicts arise and I'll try to make some time for you.
@@SirajRaval, indeed I see that you have made that choice, as well as other choices. Hopefully you'll be able to make use of the project to fix some of your repositories. Feel free to keep an open line of communication (eg. open an Issue if problems arise), and I'll be happy to help problem solve with you.
Does not look like this contains the tutorial.
Myths and realities of Data Science,AI,ML and DL : th-cam.com/video/AhIYKIvOT_o/w-d-xo.html
Despite all the sadness and disappointment regarding his misdeeds I can't resist saying the guy is talented in conveying interesting content in such a captivating way that he keeps me watching
yeah. he's got talent at least
Dude, THANK YOU. This is a great explanation! But also, the AlphaFold code available says it can't be used to predict structures outside of the CASP13 dataset. Does this mean you actually can't input your own sequence of interest, or just that accuracy cannot be ensured with other sequences?
I think since they did not publish how to extract the features ; it should not work properly with other proteins
Can you use this code to predict new protein structures? Where do you put the file with the sequence you want to predicted? Can you visualize using PyMOL?
Have you found an answer to this?
@@vitokonte no
I think you might be the only guy I've seen over 13 that can actually look good with blue hair.
(alphafold_venv) (base) jisna@ccc-CELSIUS-W570power:~$ pip install -r alphafold_casp13/requirements.txt
Could not open requirements file: [Errno 2] No such file or directory: 'alphafold_casp13/requirements.txt'
I am getting this error..
same here
You need to download the whole folder called "deep-mind research" from github to your terminal directory then run the code
How come there's always positive feedback when you just release a video and days later the other way around? Hmmm...
You have balls Siraj I give you that
Ur previous video got so much hate yet so many likes in this video. Your audience is at war with each other.
Watching this as a 50 year old GenXer...confirms that I totally dont give a fuck that young people didn't learn cursive. Ya'll know code. And code jargon. Communication tools orders of magnitude more powerful.
I run AlphaFold, but an error occurred.
It is AttributeError: module 'tensorflow' has no attribute 'compact'
.
Does anyone have the same error?
check with the github alphafold if you have the correct tensorflow version
4:25 protein sequence not genomic sequence
He doesn't know what he is talking about...
6:29 when you install pirated software and your antivirus starts complaining.
The only bioinformatics video I've seen that's entertaining.🤩😂 I'm sold.😅
Siraj, we cannot use our own sequence to predict the structures using AlphaFold, as per the GitHub website data and the alphafold website. So, how you did it?
Where is the tutorial? Waste of time!!!!!!!!!!!!!!
Finally after a while something worthy...
Don't care wht PPL say anymore continue this amazing channel.... This video is interesting
Siraj is back stronger and assertively than ever. AWESOME!
what if you accepted you don't know machine learning and AI you talk so much about.
Go on to take a break and then coming back after a year or so with strong knowledge of concepts.
just a recommend though i know u r a TH-camr and not a AI teacher in any way.
Wow, I wonder how the protein folding structre is for covid-19 could you upload a video on that!
Sorry this is late, but covid-19 is not a protein. It contains proteins, but it much more than just that.
@@nukegundieAlright thank you, I noticed that now, it has been 2 years lol
El tutorial empieza en el minuto 6:32
This is great Siraj, I have always loved your videos so keep up the hard work
Every time i try to install the code it tells me invalid syntax on venv
This is a huge breakthrough in biology.
But saying that it will replace biochemists is like saying that a stat. analysis software will replace mathematicians. AI will play a huge role in future scientific research but that doesn’t mean that it will make biochemists useless.
can you do the same for alphafold 2 or they follow the same steps?
Nice tutorial.
Next up: an actual tutorial
Hello world
Anyone knows where i can retrive this video of this part 3:40 (protein folding and sequence space)?
It will be useful for my seminar :) !
Love the content in last 2-3 videos
It sounds like Alphafold doesn't take into account post-translational modifications and chaperones.
and protein complexes if I remember that right
These are getting better
So glad you are making a comeback.
I'm glad that you didn't stop cause of the misdeed you got caught on
keep being honest or at least professional. you're actually good
we need an updated video
You are my super hero inspiration! Keep going!
Glad to see your mojo back!
Siraj, where should I input my custom amino acid sequence for predictions?
Vito Konte, did you figure out how to do this ?
Siraj now have they solved the protein folding problem or they are just bluffing
They've not solved it, at CASP they correctly predicted 25 out of 43, and also they don't really know what's going on. They are just using similar structures to predict.
@@j0nthie its like teaching a baby with neural network and huge memory spaces..am u ryt
I'm guessing Alphafold has trouble predicting intrinsically disordered proteins.
As always amazing Work by DeepMind. Thanks for sharing this :)
Sorry, but the original Alphafold implementation is a dead-end. Some others were made:
1- ProSPr (github.com/dellacortelab/prospr)
2- Alphafold_pytorch (github.com/Urinx/alphafold_pytorch)
A paper on the first is in bioRxiv since November. I would rather focus on this one.
umm.......
Learn healthcare machine learning guide pls
This video is great! I don't understand the dislikes
Rosetta has been on it for quite long, now trRosetta works like charm with AI
Hey Siraj, can you make a video on the Bayesian neural network? it's a very underrated topic in the field of machine learning but I feel like the intersection of the graphical model, neural network and Bayesian statistics will the next breakthrough
Amazing video, nicely done bro. Great content.
Hii Siraj, recently i am checking your video to understand about #auto image captioning and i am not able to found any video on this topic so if you are able to make video in this topic that will be great. BY the way loves video
th-cam.com/video/ADnD7Tlw2OY/w-d-xo.html
Someone could load a RR file in CMView? I have a format error message :/
Thank you!
So glad you’re back man, the internet sucks without you
Brilliant video!
th-cam.com/video/ADnD7Tlw2OY/w-d-xo.html
Can que simulate simple cells? If yes, can we simulate simple multicelular organisms like a worm? If yes can we simulate complex multicellular organism like a fish and it's brain? Or a human brain?????
No not yet, but when we can, the game will change big time. I’m betting 2 years to simulate a simple cell
@@SirajRaval we are all made of cells right? So when it happens we would be able to simulate a entire human being and discover the secrets of the brain and Biological inteligence???
Dev do futuro , seems like it should be that simple, right? I learned how to cook an egg, now I know how to cook all dishes that contain eggs, I learned how to hold, tune, and pluck a guitar, hold on to your face Eddie Van Halen, I’m gonna out-shred you tomorrow.
Weirdly, we can’t yet jump from fundamental principles to mastery that quickly, but soon Alpha-universe will.
If a step takes a month, and it takes 50 exponential steps to get to 100 trillion, we should be on track to get to human intelligence this decade and all intelligence in 25 years, but not tomorrow or even in the first quarter of this amazing year.
www.pnas.org/content/early/2020/01/07/1910837117
Looks like everyone is getting a like 🙄
Very informative by the way . Nice seeing the real Sitaj back 🙂
@Patrick Kuhlmann haha!
@Patrick Kuhlmann ya signs of being unlucky 😅
lol I thought in the beginning he was talking to the plant xD
well now that you moved i learned something as well never pretend to be someone who your not.. be true to ones self no matter what
Can this thing be done in windows 10?
source alphafold_venv/bin/activate i get error as source is not recognised command, some basic mistake can some one help
Did you try
$ chmod u+x alphafold_venv/bin/activate
? before the source command?
That's so epic 😍
Can we know about the protein and its working we get from this code? What is the real purpose implementation of this?
Knock knock
Whose there?
I am the Protein folding problem, cause it's still not solved.
Mind the difference between structure prediction and structure determination.
Superb as Always!!!
Siraj, there's a Git Issue that may help categorize which of your repositories can be fixed automatically and which that will require human intervention...
github.com/git-utilities/fix_logs/issues/1#issuecomment-579574567
... the default `config.json` has also been updated with those repositories that are easiest to fix via script.
Let me know if questions on resolving merge conflicts arise and I'll try to make some time for you.
I replied in the github issue
@@SirajRaval, indeed I see that you have made that choice, as well as other choices.
Hopefully you'll be able to make use of the project to fix some of your repositories. Feel free to keep an open line of communication (eg. open an Issue if problems arise), and I'll be happy to help problem solve with you.
siraj brother sorry to ask this as your old student
can you suggest a rigid roadmap to learn Tensorflow Lite
and deploy them in application
Wow!.. This is very interesting...Thank you! :)
Your subscriber count is finally increasing!!
Really? I see always 691.000 subscribers. Which is in fact extremely weird!
Nice video!
Can you please tell how to install alphafold into pc or how can we use alphafold_casp13 codes.
Pls tell me
That's incredible Siraj!
Thanks ♥
Now it is open source make video on it.
Captian Siraj ;) Like you understood
Seriously anyone knows why he always has 691,000 subscribers? I guess he made a deal with TH-cam or something?
*692.000
So great full for your contribution and dedication for democratising artificial intelligence and demystifying it.
update it!
Wow, thanks Siraj, this is truly amazing. When will you apply these techniques to crack the human genome code?
why dislikes?
I gotta come up with new comments besides "You are so awesome, on so many levels!" Thank you!
very interesting, thank you
Is there anything we can't do with Python? L.oL. No - There is not.
very nice Video
Stop moving your hands so much.
Code chor
Why do you always have 691000 subscribers Siraj? Tell the truth, what's going on? I still believe in you but this is fishy.
Omg 🤦♂️ TH-cam changed there system so it doesn’t show the exact sub count. It rounds it.
@@aayushmitra121 Ok, it just seemed shady. You never know after what he has done....
Alejandro Ruiz ya I guess I get where you are coming from
Thank you
Hey man, umm can like we become friends and stuff. You're a funny dude, plus Ummm teach me oh great and wise
kidding :)
...like the ice cream cone rotator... lol.
illuminati ?
Hello Siraj, good work, i would like to be my tutor, how to contact you, i sent you messages on Facebook and Instagram, without reply