neighbor tree construction: distance matrix not symmetric? - (Oct/24/2006 )

Dear friends,

I have a set of 164 closely related GPCR's of which i wanted to construct a tree using neighbor.

I intially took the sequences through 100 bootstraps and calculated the protein-distance matrices using protdist for the 100 datasets. However as i took the protdist output for neighbor the following error showed up.

Data set # 4:
ERROR: distance matrix is not symmetric:
(135,112) element and (112,135) element are unequal.
They are 4.544564 and 4.544564, respectively.
Is it a distance matrix?

and so i could not obtain the 100 different trees as the file stopped at dataset # 4 with the above error.....

i would be glad if you people can let me know what might be happening.

thanks in advance.....

-string-

QUOTE (string @ Oct 24 2006, 10:38 AM)

Hi friends,

I managed to find out what the problem is....

my infile (a protdist output) had matrix values that had been concatenated because they ended being large values. I manually edited them by introducing a blank and now it works fine.

The phylip authors might want to take this for a note and see that blanks are introduced when distances become too big.

till then....have a good day!!!

-string-

Is Phylip still in development?

you could alway try muscle as it writes a dendrogram to.

-perlmunky-