bug in path - bio3d-nma-muscle invocation

[jay]

Sir,

In bio3d/windows, muscle command in NMA invokes the following error : looks like a bug in the '-matrix' arg of muscle command :

exactly to say, the forward slash instead of the backward in the path of custom.mat file...

because muscle is implicitly invoked during mode generation, user does not have any control over it.

all the commands before this were successful and 'pdbs' object is created.

pl. help me resolve it ...Thank you.

> modes <- aanma(pdbs, rtb=TRUE)
MUSCLE v3.8.31 by Robert C. Edgar
Error in (function (aln, id = NULL, profile = NULL, exefile = "muscle",  : 
  An error occurred while running command
 'muscle'
In addition: Warning messages:
1: running command 'C:\windows\system32\cmd.exe /c "muscle -in C:\Users\SHREES~1\AppData\Local\Temp\RtmpY5O4rF\file17506b4e4462 -out C:\Users\SHREES~1\AppData\Local\Temp\RtmpY5O4rF\file17503749554b -matrix C:/Users/Shree Sriram/Documents/R/win-library/3.4/bio3d/matrices/custom.mat -gapopen -3.0  -gapextend -0.5 -center 0.0  -seqtype protein"' had status 1 
2: In shell(shQuote(cmd), ignore.stderr = !verbose, ignore.stdout = !verbose) :
  '"muscle -in C:\Users\SHREES~1\AppData\Local\Temp\RtmpY5O4rF\file17506b4e4462 -out C:\Users\SHREES~1\AppData\Local\Temp\RtmpY5O4rF\file17503749554b -matrix C:/Users/Shree Sriram/Documents/R/win-library/3.4/bio3d/matrices/custom.mat -gapopen -3.0  -gapextend -0.5 -center 0.0  -seqtype protein"' execution failed with error code 1
> 

[Xinqiu Yao]

Hi,

I've tested the aanma.pdbs() with my WIndows box but didn't get any errors. I am using RStudio 1.0.143 with R 3.4.1, bio3d 2.3-2, and Muscle 3.8.31.

Could you update all your software packages to the latest version and try it again? Or, you can tell me your working environment and see how I can reproduce your errors.

[jay]

Respected Sir,

Thank you for your kind reply.

I found the bug and fixed it. It was the space in the windows user name...I created a user name without any spaces and got it running.

To give a short introduction about my work, I have 16 bacterial enzyme sequences of length ~390 aa representing 7 micro-organisms. I am interested in analyzing the normal modes of the enzyme sequences by way of 'aanma' (bio3d).

First I tried doing all-atom NMA with the 'RTB' mode set to true on my linux laptop and could not get it done due to the RAM space constraints. I tried the same job on a 64-bit windows laptop with 8GB RAM / intel core i5 / 2.7GHz processor speed and could run 16 sequences in around 30 minutes.

Now, I just wanted to know whether the same job is feasible on the 64-bit windows laptop (HP Elitebook 8470p) without the 'RTB' approximation ??? By how much time and RAM space can I expect the the job to increase when RTB mode is set to 'FALSE' . And with the 8GB RAM, can we expect the job to proceed to completion without crashing the OS ? Please advise on the matter as I would like to look at the fluctuations without the 'RTB ' approximation but apprehensive about any crash ( due to inadequate RAM space because of the increase in eigen vectors) . . .

Expecting your valuable opinion on the matter, Thanking you, Jayanthi. J.

[Xinqiu Yao]

Hi,

rtb=FALSE would be slower. There is a progress bar on the bottom when you run the aanma() function, with which you can roughly estimate how much it will take to finish.

I think the memory usage between rtb TRUE and FALSE should be the same. So, if you can run with rtb=TRUE, it should be okay to go with rtb=FALSE.

Let me know if you run into any troubles.