Hello Nela,
Can you provide further information regarding the parallel cluster
architecture for which you have compiled Mauve?  It is possible that on
some exotic architectures the software does not operate properly as it
performs bitwise operations and stores the results in files, so
mismatches in endian-ness can be a problem.  

Since you built the code yourself, I'll also need to know exactly which
revision you compiled before I can help you solve this problem.  Was it
a source code snapshot?  The subversion trunk?  If so, which revision?

Recent versions of the codebase have portions that are parallelized with
OpenMP.  That offers some degree of parallelism, but several time
consuming portions of the alignment algorithm still execute in serial.  

Since this conversation is veering in a very technical (e.g. software
development) direction, I would prefer to keep it off-list for now (and
spare people's inboxes our jargon). Once we have some resolution later
on it would be good to post it for the record.

Regards,
-Aaron


On Tue, 2009-04-21 at 16:51 -0400, Nela Zavaljevski [Contractor] wrote:
> Hi,
> 
>  
> 
> Our group compiled MAUVE on Red Hat Linux using instructions provided
> on MAUVE’s web page.   We wanted to reproduce the results for 5
> Salmonella genomes from ERIC’s web page.  We used both mauveAligner
> and progressiveMauve with the default options, but could not obtain
> the results from the web page.  Moreover, progressiveMauve crashed for
> the default setup with the following error message:
> 
>  
> 
> cga doesn't fit
> 
> cga.RightEnd(cga_seq) 292746
> 
> RightEnd(my_seq): 5133713
> 
> cga.LeftEnd(cga_seq) 292724
> 
> LeftEnd(my_seq): 4836644
> 
> cga.AlignmentLength(): 23
> 
> AlignmentLength(): 292727
> 
> cga doesn't fit
> 
> cga.RightEnd(cga_seq) 293601
> 
> RightEnd(my_seq): 5133713
> 
> cga.LeftEnd(cga_seq) 293579
> 
> LeftEnd(my_seq): 4836644
> 
> cga.AlignmentLength(): 23
> 
> AlignmentLength(): 292727
> 
> progressiveMauve: 
> /local/software/mauve/include/boost-1_34_1/boost/dynamic_bitset/dynamic_bitset.hpp:952:
>  Unknown( ) : Assertion `pos < m_num_bits' failed.
> 
> Aborted
> 
>  
> 
>  
> 
> We tried to change the breakpoint penalty as suggested in Example 7 of
> the User’s manual and we were able to get some results. However,
> results were frequently irreproducible:  in some runs the program
> crashed for the given value of the weight parameter, with an error
> message similar to the previous one, while in other runs with
> identical input the program worked fine.   Has anyone had similar
> experience and what could be the reason for such behavior?  In
> general, we are not sure how to select parameters for progressiveMauve
> when the default setup does not work and would appreciate any advice
> in that matter.
> 
>  
> 
> Thank you very much,
> 
>  
> 
> Nela Zavaljevski
> 
> 
> ------------------------------------------------------------------------------
> Stay on top of everything new and different, both inside and 
> around Java (TM) technology - register by April 22, and save
> $200 on the JavaOne (SM) conference, June 2-5, 2009, San Francisco.
> 300 plus technical and hands-on sessions. Register today. 
> Use priority code J9JMT32. http://p.sf.net/sfu/p
> _______________________________________________ Mauve-users mailing list 
> [email protected] 
> https://lists.sourceforge.net/lists/listinfo/mauve-users


------------------------------------------------------------------------------
Register Now & Save for Velocity, the Web Performance & Operations 
Conference from O'Reilly Media. Velocity features a full day of 
expert-led, hands-on workshops and two days of sessions from industry 
leaders in dedicated Performance & Operations tracks. Use code vel09scf 
and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf
_______________________________________________
Mauve-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mauve-users

Reply via email to