Hi Sergei,

I have a questions for you.

I would like to do NeighborJoin from a saved partition but I have problems:
for example, 9 sequences with 5000 sites. I select 1-2000 sites and make a partition. I save that partition in sequential or phylip format. Then, I load the partition, it's fine.
Now, I want to make NeighborJoin to the partition so I select all sites (1-2000) and make a new partition to do NJ, when the program try to do NJ an error occured: "Invalid distance matrix passed to NJ. Matrix written onto messages.log Current BL Command:njm=(distanceMatrix>methodIndex)>=_NUMBER_OF _SEQUENCES Current operation / job has been terminated".

If I use the same sequences and sites with other phylip format (there isn´t in program, continue all sites, see below), there isn't problems. For example:

Problems, phylip sequential format from Hyphy
seq00001   AGGGCACGCT GGGTAGCTGG GGCGGCTTGC GGACAGCTCA AGACACAATG
          CGCATCCACT AAACTGAGTA TGTAGGAGCC TCCTGGGAGA CGGGCCCCGC
          GCTAGTCGGT TGTTCTATTA AAGCCCAGTG AGTTACTTTC GACTACGAGA

No problems, phylip format from me (not hyphy)
seq00001   AGGGCACGCTGGGTAGCTGGGGCGGCTTGCGGACAGCTCAAGACACAATGCGCATCCACTAAACTGAGTATGTAGGAGCCTCCTGGGAGACGGGCCCCGCGCTAGTCGGTTGTTCTATTAAAGCCCAGTGAGTTACTTTCGACTACGAGA

What's your opinion?
Is it a problem of files format?, translate the format Hyphy phylip to my phylip format is too many time for me (long sequences).
Do you know what can I do?
Thanks you!!

Cheers.
Miguel.

 ::)

Dear Miguel,

This sounds like a bug. Could you send me the file you were using by e-mail, along with the sequence of steps (be as detailed as possible) that result in the error message and I will try to recreate and troubleshoot the issue.

Cheers,
Sergei

Hi Sergei,

I'm Sorry, don't worry, I could solve the problem.
I found a mistake in my sequences. The sequences have too many mutations so the NJ isn't possible. With few mutations the NJ is fine. Good Hyphy!!

Cheers,
Miguel.

::)

Dear Miguel,

I have seen this happen before; should have occurred to me as well.
Glad you found the problem.

Cheers,
Sergei