[BioC] running time of snapCGH processCGH and runDNAcopy

Sean Davis sdavis2 at mail.nih.gov
Fri Oct 5 23:03:26 CEST 2007


Henrik Bengtsson wrote:
> On 10/5/07, Sean Davis <sdavis2 at mail.nih.gov> wrote:
>> jhs1jjm at leeds.ac.uk wrote:
>>> Hi Caimiao,
>>>
>>> For me, the DNAcopy segmentation function ran quicker in the actually
>>> DNAcopy package than it did in the snapCGH package, not sure why but I
>>> do know they made some speed improvements in the CBS algorithm that may
>>> not have been implemented in snapCGH. It may be worth investigating
>>> using the actual DNAcopy package, especially if snapCGH is crashing on
>>> you. If anyone feels this is wrong feel to correct me, just thought I
>>> should contribute as I've made heavy use of this mailing list myself!
>> Hi, John.
>>
>> Just to clarify, snapCGH uses the DNAcopy package.  They are not
>> different implementations.

Looking at the runDNAcopy code, it does the smooth.CNA step as well as
running DNAcopy::segment(); that probably explains the slowdown when
using runDNAcopy.

Sean



More information about the Bioconductor mailing list