How to use AlphaFold2 for Wet Lab Researchers

Поділитися
Вставка
  • Опубліковано 26 лип 2021
  • My effort to explain how to use the Alphafold2 algorithm to predict the structure of a protein without having a lot of bioinformatics knowledge.
    I am mostly a wet lab researcher and I know that a lot of us are far from confident in our bioinformatics skills. That is why it was important to me to publish this video, so more people can benefit from this amazing technology.
    Links:
    Blog post on Alphafold2
    deepmind.com/blog/article/put...
    Alphafold Protein Structure Database on ebi website
    alphafold.ebi.ac.uk/
    Alphafold GitHub page to download the code
    github.com/deepmind/alphafold
    alphafold Colab page to run it on the cloud
    colab.research.google.com/git...
    Colab Pro to have more computational resources
    colab.research.google.com/signup

КОМЕНТАРІ • 17

  • @orhunakengin7072
    @orhunakengin7072 Рік тому

    Such an underrated video, thanks for explanation!

  • @lunanaw5867
    @lunanaw5867 2 роки тому

    Thanks! Great explanation.

  • @maxhuber1361
    @maxhuber1361 Рік тому

    thank you so much!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! I am waiting for the continuation

  • @Maria-gb8bv
    @Maria-gb8bv Рік тому

    Thank you!!

  • @RavindraThakkar369
    @RavindraThakkar369 2 роки тому

    I have two peptides with 17 residues each, one is head to tail cyclic peptide and the other is stapled peptide. Does alphafold2 work with chemically modified peptides?

  • @dandeelyonn
    @dandeelyonn 2 роки тому +3

    Hey Reza, could you please make a tutorial for using rosettafold please? Maybe that's possible to run locally since the requirements are lower?

    • @rezarezaei4868
      @rezarezaei4868  2 роки тому +1

      I think there is also a full online version of rosettafold that I can record a video on it and for that I don't think there is any need for a local version. For alphafold the local version uses the full database and the online one uses a smaller database. So, the local's result is more accurate in some cases.

    • @dandeelyonn
      @dandeelyonn 2 роки тому

      @@rezarezaei4868 hi Reza,
      Yes I did some digging yesterday and found that the Baker lab website has an online submission point so it is easy enough to look at individual proteins- however I am unsure of how to build complexes (I heard that if you just add a flexible linker and repeat the monomer N Times that can work, so might try that with the server submission)

  • @shrutiagrawal8026
    @shrutiagrawal8026 6 місяців тому

    at the time of installing the database is shows error "Error: aria2c could not be found. Please install aria2c (sudo apt install aria2)." in spite of having aria2 installed.

  • @deby6293
    @deby6293 2 роки тому

    Hello Reza what to do in case you want to predict a dimer?

    • @rezarezaei4868
      @rezarezaei4868  2 роки тому

      Hi, actually I haven't tested this myself but I read that if you put a monomer first and then a hundred X and then the sequence of the second polypeptide, you can predict the dimer structure. But I'm not sure how accurate that would be. There have been some interesting results using this method! However, the current version does not natively support dimers.

  • @jamesbean9396
    @jamesbean9396 2 роки тому +4

    The champion of computer illiterate wet lab researchers xD Cheers mate, great video!

  • @premlatamajhi7059
    @premlatamajhi7059 Рік тому

    U should have made it clear for those who don't have any dependencies what should they do

  • @omidmokhtari4606
    @omidmokhtari4606 Рік тому

    "damet garm" from Stockholm xD

  • @marcelosilvafolhasdamas1329
    @marcelosilvafolhasdamas1329 2 роки тому

    What are your computer parameters? The minimum requirements for running it?

    • @rezarezaei4868
      @rezarezaei4868  2 роки тому

      I think the minimum is at least 8 gb of gpu and 32gb of ram. Also you need around 3Tb of fast storage for having a local copy of all the libraries.