[chimerax-users] alphafold predict stops after MSA step (google colab pro)

Tom Goddard goddard at sonic.net
Wed Feb 23 21:49:45 PST 2022


Hi Amith,

I've seen Google Colab disconnect if your computer sleeps for an hour or two.  So if you are running ChimeraX AlphaFold prediction which uses Google Colab you may need to change your computer settings so it does not sleep for a few hours.

Google Colab documentation gives very little information about when you will be disconnected.  I thought as long as a computation is running it would not disconnect but I have not seen any Google documentation that says what will happen.

	Tom

> On Feb 23, 2022, at 8:13 PM, Rangarajan, Amith via ChimeraX-users <chimerax-users at cgl.ucsf.edu> wrote:
> 
> This Message Is From an External Sender
> This message came from outside your organization.
> Hi,
> 
> I am new to using alphafold for prediction, i tried today with a reasonable size model ( 400 AA ) of a kinase domain. i opted to use 'high RAM " available on colab pro and after running for an few hours , the job stops without any error, the colab notebook had disconnected due to inactivity and  it did not generate any models.
> 
> please share with me your experience of using alphafold from within chimerax , helpful tips to get the job to finish .
> here is the output ( output on google colab):
> Sequence length 401
> Have Colab GPU runtime
> Installing HMMER for computing sequence alignments
> Installing matplotlib to plot sequence alignment coverage
> Installing AlphaFold
> Installing OpenMM for structure energy minimization
> Finding fastest mirror for sequence databases using united states
> Searching sequence databases (147 Gbytes).
> Search will take 29 minutes or more.
> Searching uniref90 sequence database, 59 Gbytes
>  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
>  31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
> Searching smallbfd sequence database, 17 Gbytes
>  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
> Searching mgnify sequence database, 71 Gbytes
>  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
>  31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
>  61 62 63 64 65 66 67 68 69 70 71
> Computing multiple sequence alignment
> 327620 similar sequences found (311112 uniref90, 16008 smallbfd, 500 mgnify)
> <image.png>
> Computing structures using 5 AlphaFold parameter sets:
>  model_1 
> 
> 
> thanks a lot again for all your support,
> 
> best
> amith
> _______________________________________________
> ChimeraX-users mailing list
> ChimeraX-users at cgl.ucsf.edu <mailto:ChimeraX-users at cgl.ucsf.edu>
> Manage subscription:
> https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users <https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.rbvi.ucsf.edu/pipermail/chimerax-users/attachments/20220223/7b14e6cf/attachment.html>


More information about the ChimeraX-users mailing list