CODEML exits with no error message.

[Bin He]

Dear all,

I've been working with this problem for quite a while and still no clue what is wrong. I'm basically trying to replicate the nef (HIV-2) analysis posted in the paml demonstration ppt (by Joseph Bielawski

). I've downloaded the 44 alleles from pubmed, aligned them after translating them to amino acid sequences, back-translated them to cDNA alignment using pal2nal.pl.  I used nj method to infer the tree and store it in a text file.

My ctl file looks like the following:

seqfile = nef_Padua_2003.codon.phy

outfile = results.txt

treefile = nef_Padua_2003.trees

verbose = 1

CodonFreq = 2

NSsites = 0 1 2 3 7 8

fix_omega = 0

runmode = 0

fix_kappa = 0

icode = 0

seqtype = 1

omega = 5

noisy = 9

kappa = 2

model = 0

But after running codeml, the program finishes "counting patterns", then report the following before quitting without any error message:

     7920 bytes for distance

   316224 bytes for conP

    44064 bytes for fhK

  5000000 bytes for space

Model 0: one-ratio

Species gi|30026968|emb|AJ344409.1|?

Any suggestion would be most appreciated!

Thanks

Bin

post-doc fellow

Harvard University, FAS Center for Systems Biology

bi...@fas.harvard.edu

[anavale...@gmail.com]

Hi,
It also happens to me when codeml does not recognize the name of that given species or taxon from the tree or alignment files. I would check where you have that "gi|30026968|emb|AJ344409.1|", and to change it for a proper name. Names of species in tree and alignment files must be exactly the same.
I hope it helps,
Ana

[Bin He]

Thanks Ana! I've since moved to my real data, for which I have had proper names. And for those I've not encountered this problem. I will follow your advice and check on my toy data again to see if that's indeed the cause. Thanks again!

Bin

[HZ]

I had similar problem and it took me a while to figure out that the last line in the output is actually an error message! It has the form: Species XXX ? By reading the threads in the forum, I realize this happened to many users. How about making the error/warning/reminder message more prominent, dear Paml development team ? :-)

[František Zedek]

I think there have to be at least two spaces between a name of the sequence and the sequence itself, otherwise PAML thinks that the sequence is a part of the taxon name.

[Ziheng]

The seq name is causing problems.
I'll write a longer error message.
Ziheng