Unable to run pdflatex

Question

[jean.charles.cancel]

Hello everyone,

I am new to bioinformatics and struggling with rstudio already. I am trying to perform a Multiple Sequence Alignment using the msa.package

I am currently having 2 kind of issues.

---

• First of, after using the msaClustalOmega(sample) everything is working fine then I run the print(sample, show = "complete")

But then when I want to use the msaprettyprint() function here is what is displayed.

> msaPrettyPrint(nifHomega)
File nifHomega.tex exists. Overwrite? (y/N)
1: y
Error in texi2dvi(texfile, quiet = !verbose, pdf = identical(output, "pdf"),  : 
  unable to run pdflatex on 'nifHomega.tex'
LaTeX errors:
! TeX capacity exceeded, sorry [save size=80000].
\inf@@get ...@ {#5} \else \def \fourth@ {99999999}
                                                   \fi \fi \def \fifth@ {#5}...
l.25 ...de}{/tmp/RtmpldGSDm/seq17866bb458cf.fasta}

!  ==> Fatal error occurred, no output PDF file produced!

I already tried many different things and I can't manage to find a solution. I really need the prettyprint() as I want to see find a consensus sequence in my alignment.

If you have any solution, or any other mean to identify a consensus sequence of 10-15bp from ;ore than 18000 sequences please tell me.

---

• My second issue is that the msaClustalW() is taking ages to run on my .fasta document. Until now it's been working for 8h and still won't be finished. Any idea on what's happening there ? The ClustalOmega is working fine but not the ClustalW...

Many many thanks in advance. Bests,

[antonioggsousa]

Hi,

I never used that package, so I'm not familiar with it. From the error message - ! TeX capacity exceeded, sorry [save size=80000]. - it seems that you cannot run pdflatex because you run out of RAM memory or the memory exceed by the program. Therefore, the memory that you've is not enough to produce the file that you want. Based on the number of sequences that you have that makes sense because it is a quite few thousands of sequences. Actually, I don't understand how you'll try to find manually 10-15 bp consensus sequences in 18 K seqs. I don't think this is feasible.

António

Answer 1

[Istvan Albert]

It sounds that in both cases the datasize is too large to be rendered.

In general I would recommend to run the aligner tools from command line then find ways to visualize the results.

These R wrappers can often add large burdens that can make a run fail.