Partition alignment BEAST1.10 vs BEAST2.5.1

[Eve]

Hi all,

I am having a problem loading my data in Beauti2.5.1. Googling the problem did not give me any indication on how to solve the issue, so I decided to post it here.

I want to run an analysis in BEAST2.5.1 using data of six genes. Genes 1 and 2 are not partitioned by codon. Genes 3-6 are partitioned by codon, but some of these codons are merged into a single partition (the first codon positions of genes 4-6, the second codon positions of genes 4-6, and the third codon positions of genes 4 & 6). In my alignments, I want to keep all positions of each gene together (instead of physically separating and ordering the codons). I therefore added an extra data block to my nexus file which indicates the partitions I want to use:

begin sets;
charset gene1=1-1258;
charset gene2=1259-2368;
charset gene3_1=2369-3112\3;
charset gene3_2=2370-3112\3;
charset gene3_3=2371-3112\3;
charset gene4_1_gene5_1_gene6_1=3113-4470\3, 4471-5366\3, 5367-6380\3;
charset gene4_2_gene5_2_gene6_2=3114-4470\3,4472-5366\3,5368-6380\3;
charset gene4_3_gene6_3=3115-4470\3, 5369-6380\3;
charset gene5_3=4473-5366\3;
end;

This loads perfectly well in Beauti1.10 (and any other BEAST1 versions I tried). However, Beauti2.5.0 and 2.5.1 consistently give the following error message:

Warning: The following partitions overlap:
gene1 overlaps with gene4_1_gene5_1_gene6_1
gene1 overlaps with gene4_3_gene6_3
The first thing you might want to do is delete some of these partitions.

It is however obvious from my nexus data block, as well as the fact that Beauti1 reads the file, that these error message is wrong. Gene 1 does not overlap with any of the other partitions. Some playing around with my data indicated to me that the problem arises when merging the different codons (i.e. when creating charset gene4_1_gene5_1_gene6_1, charset gene4_2_gene5_2_gene6_2 or charset gene4_3_gene6_3). Beauit2.5.1 correctly reads the file when the following partition scheme is given:

begin sets;
charset gene1=1-1258;
charset gene2=1259-2368;
charset gene3_1=2369-3112\3;
charset gene3_2=2370-3112\3;
charset gene3_3=2371-3112\3;
charset gene4_1=3113-4470\3;
charset gene4_2=3114-4470\3
charset gene4_3=3115-4470\3;
charset gene5_1=4471-5366\3;
charset gene5_2=4472-5366\3;
charset gene5_3=4473-5366\3;
charset gene6_1=5367-6380\3;
charset gene6_2=5368-6380\3;
charset gene6_3=5369-6380\3;
end;

Do you have any recommendations on how to solve this issue? I really want to avoid sorting through all my data as this is yet another step that could introduce errors.
In case it is of any use: please find attached the nexus file (test.nex) that is giving me the error.

Many thanks already.

All the best,
Eve

[Walter Xie]

Hi Eve,

BEAST 2 has the different NEXUS parser to BEAST 1, which is using slightly different grammar comparing to BEAST 1. This “charset” is the case. BEAST 2 uses white space to split multiple sections in “charset”, not comma. See the example below. I also attached the screen shot of loading your NEXUS file after the delimiter “,” is replaced to “ ”.

 

charset firsthalf = 1-449;

charset secondhalf = 450-898;

charset third = 1-457\3 662-896\3;

 

Cheers,

Walter