Parallel computing for Cobra Toolbox

[stanley paul]

Hi...

I installed my Cobra Toolbox on 32 bit, windows 7 platform on 4Gb ram....along with Tomlab

when I am trying to run double gene deletion at one point ....Matlab is not unable to run further....but the same model once completed when I used in.... Mac operating system....

I think the problem is because of cores that are being used by Matlab... even though I used Parallel computing for connecting all my cores in my local computer..Its not increasing..its capacity/ power to solve....

Hw can I run Cobra Toolbox by using parallel computing...

1)  Can I use 2 Pc power for Cobra Toolbox?

2)  do I have to install Cobra toolbox on both systems along with Tomlab or not?

Please give me suggestions ....

Regards

[eve kuhestani]

hi paul

as i remember my supervisor discussed about same problem coz it seems i will need same power to solve my own question working in my thesis..you have 2 ways i think..first use more capable system with higher ram memory or use several systems connected to each other that every one should have same program and same basics

this  was my supervisor's suggestions 

[stanley paul]

Hi eve....

Thats what parallel computing says.....I used it in my cluster(about 64 cores).....system of my institute but it seems to be connected while we are running the models of Cobra toolbox...its not taking the power of all 64 cores... ....and ram also...having same problem......is it is the problem of Matlab or Cobra Toolbox???.....I am not getting.....eve...

Regards 

M.L. Stanley Paul, M.S.Pharm.

Research Block - A, 

Room no - 208 
Department of Pharmacoinformatics,

National Institute of Pharmaceutical Education and Research(NIPER),
S.A.S. Nagar, Mohali,
Punjab :-160062,
INDIA.

Cell. no.     : +91 - 9646981658

E-mail        : stanleyp...@gmail.com

                   mlst...@niper.ac.in

[Carlos Javier Almeciga]

it is probable that your mac is running a 64 bits matlab version. I had the same results (PC vs Mac) and that was the only difference. In fact, my mac was 2.2 Gb vs a 4 Gb PC. In addition, matlab only use one processor. That's a limitation of the software.

Javier

Enviado desde un doniBerry®
Sent from my mobile device

---

From: stanley paul <stanleyp...@gmail.com>

Sender: cobra-...@googlegroups.com

Date: Tue, 25 Sep 2012 04:47:52 -0700 (PDT)

To: <cobra-...@googlegroups.com>

ReplyTo: cobra-...@googlegroups.com

Subject: Parallel computing for Cobra Toolbox

[eve kuhestani]

in fact optional power of clusters as u used is less than theoretical power..i think u should Javier is right..softwares has their own limitations too..i  think you would check 64 bit system coz i use it on my laptop and no problem yet about processing maybe u can do it in longer time but it will answered finally 

[Lindsay Edwards]

I expect the computer scientists will have more elegant ways of doing it, but I have been able to parallelize a number of COBRA methods using the Matlab parallel computing toolbox:
http://www.mathworks.co.uk/products/parallel-computing/index.html;jsessionid=d64fc868d6aa1b8daf538071865e

I have had excellent results with Monte Carlo methods.
For example, with the ACHR sampler it was relatively straightforward to have multiple instances running at once.

Best,

Lindsay.

--
Dr Lindsay M Edwards MSc DPhil (Oxon) MSB
Lecturer
School of Biological Sciences
University of Essex
Wivenhoe Park
Colchester
Essex
CO4 3SQ

Tel: 01206 873322

[Ronan Fleming]

Carlos, Eve,

What are we supposed to write in the next grant application to support
the cobra toolbox google group, that we assisted the following
institutions (a) & (b)?

(a) http://www.google.nl/search?aq=f&sourceid=chrome&ie=UTF-8&q=Carlos+Javier+Almeciga

(b) http://www.google.nl/#hl=en&safe=off&output=search&sclient=psy-ab&q=eve+kuhestani&oq=eve+kuhestani&gs_l=hp.3...1112.1112.0.1375.1.1.0.0.0.0.99.99.1.1.0.les%3B..0.0...1c.1j2.4tJ5a8FG13k&pbx=1&bav=on.2,or.r_gc.r_pw.r_qf.&fp=58f465a66cc1be8&biw=1280&bih=658

Ronan

--
--
Mr. Ronan MT Fleming B.V.M.S. Dip. Math. Ph.D.
-----------------------------------------------------------------
Assistant Professor,
Dept. Biochemistry & Molecular Biology,
Faculty of Medicine,
University of Iceland,
&
Center for Systems Biology,
University of Iceland,
Sturlugata 8,
101 Reykjavik,
Iceland.
http://www.hi.is/~rfleming
http://systemsbiology.hi.is/
Ph: +354 618 6245
Skype: ronan.fleming
-----------------------------------------------------------------

[Carlos Javier Almeciga]

Dear Ronan,

I'm sorry for this inconvenience. Most of the time I answer from my mobile device and I have not set up my signature here. I will try to include my affiliation in future e-mails. In this sense, this is my information:

Carlos Javier Almeciga-Diaz, BPharm, PhD
Assistant Professor
Pontificia Universidad Javeriana
Bogota
Colombia

Regards,

Javier
Enviado desde un doniBerry®
Sent from my mobile device

[eve kuhestani]

Dear ronan:

i really didnt get what do you mean. i attached my academic email as affiliation to my current email.

ha_kooh...@ms.tabrizu.ac.ir

i am not member of any social network.so i dont confirm the institutions you mentioned.

regards

[Ronan Fleming]

Hi Eve,
we would not be listing each individual assisted by the google group,
this is not realistic. However, we would list their institutions. It
is best in scientific correspondence to have you institutional
affiliation, as it saves detective work trying to figure out where
ms.tabrizu.ac.ir comes from.
Ronan