[ensembl-dev] Huge branch lengths in gene trees

[Sébastien Moretti]

Thanks Javier
I can skip my analyses on these branches for now.

Sébastien

> Dear Sébastien
>
> Indeed, that is an upper branch length limit set by TreeBeST. This
> family is especially difficult to resolve as it contains essentially 3
> types of genes: full length, N-terminal part only or C-terminal part
> only. As such, many pairs of genes do not overlap at all in the
> alignment. As you can imagine, it is very difficult for the phylogenetic
> reconstruction algorithm to address these families.
>
> We are working on a couple of developments that might help in those
> situations. We will let you know more if we make any progress.
>
> Javier
>
>
> On 16/10/12 15:29, Moretti Sébastien wrote:
>> Hi
>>
>> I found some gene trees that have huge branch lengths:
>> e.g. tree ENSGT00560000076605, gene ENSCJAG00000001973
>> The branch leading to that gene/leaf is  100000  long.
>> http://www.ensembl.org/Callithrix_jacchus/Gene/Compara_Tree?g=ENSCJAG00000001973;r=ACFV01174532.1:4628-45447;t=ENSCJAT00000003772
>>
>>
>> What does it mean ?
>> Is it an upper branch length limit ?
>>
>> Thanks

-- 
Sébastien Moretti
SIB Vital-IT EMBnet, Quartier Sorge - Genopode
CH-1015 Lausanne, Switzerland
Tel.: +41 (21) 692 4079/4221