EOF error?

[Monera Gam]

err reading site 7717, seq 1 group 156

error: EOF?sites read in each seq:

The above is the error for paml version 4.10.7. The phylip file itself has 7716 sites and the same analysis run on another gene that doesn't have duplications (when present, have _v0 or _v1 appended at the end of the 7 characters sequence name) doesn't give any errors. Any suggestions? Attached is the codeml.ctl file (phylip file is too big).

[Monera Gam]

Treefile

[Janet Young]

I don't know if this will fix your problem, but I did notice that your treefile is missing the first line that should specify how many species and how many trees are in the file. The tree itself should be on the second line.  

See p15 of the PAML manual, https://github.com/abacus-gene/paml/blob/master/doc/pamlDOC.pdf

If that doesn't fix it, the problem might be in your alignment file - not sure

[Monera Gam]

Tried, unfortunately it doesn't fix the problem. As I already noted, analyses run on tree, phylip and codeml.ctl files for other genes use the same format as this one but produce an output. The only difference should be the presence of sequence names like sequence_v0 and sequence_v1 together with other names missing the _v part, as it can be seen in the treefile. Maybe I'm missing something..

[Monera Gam]

[Janet Young]

interesting - because your alignment file is in "interleaved" format, I took a look at the interleaved example files provided by PAML (abglobin.aa and abglobin.nuc) - those example files doesn't seem to match the format of examples I see elsewhere (e.g. http://lorenzogatti.me/2017-ACGT-Tutorials/docs/howto_Format_Data/ ).

Sandra, or someone else in the codeml team, is PAML's understanding of phylip-interleaved format different from other definitions?  Are the numbers I see in abglobin.nuc to specify the start position of each block necessary?  Does PAML want seqnames alone in the first block? (other definitions seem to have the seqnames AND the first portion of the sequence in that first block).

Monera, your alignment DOES seem to match "standard" phylip-interleaved format, except that you have spaces at the start of the lines for the blocks after the first one, and the examples I saw do not, neither does abglobin.nuc - I don't know how much that matters.

Are you able to convert your alignment file to phylip-sequential format?  That's my next guess about what might make it work: I feel like that might be a more robust format (that's what I use myself).  But I don't know why it would have worked on your first seqfile but not your second. 

good luck!

Janet

[Sandra AC]

Hi everyone!

I have run a quick test and it seems the problem has to do with some sequences not being separated by two or more spaces (i.e., sequences that correspond to taxa which ID ends with `_v0` and `_v1`). In essence, just add a space between the taxa ID in lines 11, 12, 153-158 and the analysis should start running without the error you pointed out -- at least, this is what I see when troubleshooting your error with PAML v4.10.7!

Hope this helps!
S.

P.S.: You can find all relevant information regarding data formatting in [the PAML documentation](https://github.com/abacus-gene/paml/blob/master/doc/pamlDOC.pdf) and [the PAML Wiki, section "Sequential and Interleaved format"](https://github.com/abacus-gene/paml/wiki/Data-formatting#sequential-and-interleaved-formats). 

[Janet Young]

nice!   sorry about the red herrings from me

[Monera Gam]

Sorry for the late reply.

Thank you Sandra, that solved it!

Kind thanks to both of you for the attention given to the issue and the quick replies!