run time is fine for a dimer, but way too slow for a 700 protein fasta database in a complex sample?

49 views
Skip to first unread message

Christian Trahan

unread,
May 3, 2019, 1:29:13 PM5/3/19
to SIM-XL Group
Hi

I have ran dimers analysis on a dual cpu 3.4 GHz (24 cores) with an SSD drive and 64G of ram, and it ran like a charm. However when I try to analyze purified cross-linked complexes from ex-vivo purifications it the run time goes up to 50% within a few minutes, then it takes a week to process an additional 1% (51% in one week).

Any ideas or suggestions? Any plans on making changes to the coding so that larger dataset can be analyzed using SIM-XL in a reasonable time scale?

Cheers,

Christian
sim-xl after 1 week running.PNG

Viviane Bastos

unread,
May 3, 2019, 1:35:55 PM5/3/19
to si...@googlegroups.com
Hi Christian,

Are you considering variable modifications in your search? In my experience, including variable modifications (such as oxidation of Methionine) increases significantly the time it takes to complete the XL search. 

Cheers,

Viviane 

--
You received this message because you are subscribed to the Google Groups "SIM-XL Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simxl+un...@googlegroups.com.
Visit this group at https://groups.google.com/group/simxl.
To view this discussion on the web visit https://groups.google.com/d/msgid/simxl/df16a203-9652-4e49-b7d8-987df1c1e630%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


--
Dra. Viviane de Almeida Bastos

Laboratório de Toxinologia
Instituto Oswaldo Cruz
Fundação Oswaldo Cruz

Telefone: +55 21 25621263
Skype: viviane_abastos

Paulo C Carvalho

unread,
May 3, 2019, 2:19:00 PM5/3/19
to SIM-XL
Christian,

Do you have access to cleavable cross-links? I ask this because it would be more appropriate for the task at hand and we have a beta sim version for that.

Paulo

--

Christian Trahan

unread,
May 3, 2019, 2:46:49 PM5/3/19
to SIM-XL Group
Thanks for the response Paulo,

So in other words you guys are not planning on adding a different search algorithm for large database in order to reduce the running time? I guess it is running on exhaustive mode no matter the size of the database right?

Thanks for your input Paulo,

Best,

Christian





On Friday, 3 May 2019 14:19:00 UTC-4, Paulo C Carvalho wrote:
Christian,

Do you have access to cleavable cross-links? I ask this because it would be more appropriate for the task at hand and we have a beta sim version for that.

Paulo

On Fri, May 3, 2019, 2:29 PM Christian Trahan <Christi...@ircm.qc.ca> wrote:
Hi

I have ran dimers analysis on a dual cpu 3.4 GHz (24 cores) with an SSD drive and 64G of ram, and it ran like a charm. However when I try to analyze purified cross-linked complexes from ex-vivo purifications it the run time goes up to 50% within a few minutes, then it takes a week to process an additional 1% (51% in one week).

Any ideas or suggestions? Any plans on making changes to the coding so that larger dataset can be analyzed using SIM-XL in a reasonable time scale?

Cheers,

Christian

--
You received this message because you are subscribed to the Google Groups "SIM-XL Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to si...@googlegroups.com.

Paulo C Carvalho

unread,
May 3, 2019, 11:47:52 PM5/3/19
to SIM-XL
For complex cenários we recommend going cleavable. 

To unsubscribe from this group and stop receiving emails from it, send an email to simxl+un...@googlegroups.com.

Visit this group at https://groups.google.com/group/simxl.
To view this discussion on the web visit https://groups.google.com/d/msgid/simxl/3ce30fd2-6752-418b-893c-cb9058291a22%40googlegroups.com.

Diogo Borges

unread,
May 4, 2019, 2:07:58 AM5/4/19
to si...@googlegroups.com
Dear Christian,

In your case, we recommend to use cleavable cross-linkers. 
SIM-XL is optimized to perform searches in "target mode", that is, when is known a set of protein that likely they are interacting with each other. In this "target mode", we recommend to use a database with a number of protein sequences up to 100.

The current version of SIM-XL is not optimized to work with cleavable cross-linkers, but soon we'll release a new version with this module working and with an optimization in the target mode as well.

Best regards,
Diogo
_____________________________________

Diogo Borges Lima
Postdoctoral fellow in Computational Biology
Mass Spectrometry for Biology Unit
Institut Pasteur - CNRS USR 2000
Computational Mass Spectrometry Group - Fiocruz / Brazil

About me: @diogobor | diogobor
URL: diogobor


To view this discussion on the web visit https://groups.google.com/d/msgid/simxl/CAOVP5VO0Re7Yjn3Prz9PJY0dYF5C9SGVcJLQGszuR-3ZURwX9A%40mail.gmail.com.

Paulo C Carvalho

unread,
May 4, 2019, 1:19:40 PM5/4/19
to SIM-XL
Please let us know if you want to be a beta tester. 

To view this discussion on the web visit https://groups.google.com/d/msgid/simxl/CAKvKiVS0BMwh_AhjKt2Z38gooTzR4kpZbabA6ZGsigy8qLWixg%40mail.gmail.com.

TRAHAN Christian

unread,
May 6, 2019, 11:11:37 AM5/6/19
to si...@googlegroups.com

Hi Paul,

 

I will have to think a bit more on your offer. When does your offer expires?

 

Thanks for your responses Paul,

 

Christian

Diogo Borges

unread,
May 6, 2019, 11:14:11 AM5/6/19
to si...@googlegroups.com
Dear Christian,

We’re gonna have a initial prototype in three months. If you want to be a beta tester, tell us. 

Best regards,
Diogo 


For more options, visit https://groups.google.com/d/optout.
--

TRAHAN Christian

unread,
May 6, 2019, 11:31:32 AM5/6/19
to si...@googlegroups.com

Great! It is noted and will let you know.

 

Thanks for the heads up,

 

Christian

Paulo C Carvalho

unread,
May 6, 2019, 11:34:18 AM5/6/19
to SIM-XL
It doesn't expire.  It is a bit two early, but I think in one more month if you want to use the software we can share it with you eventhough it will be preliminary.


To view this discussion on the web visit https://groups.google.com/d/msgid/simxl/703d4ca1e4704d40a506ad508bb0363d%40MSXCHG01.ircm.priv.

For more options, visit https://groups.google.com/d/optout.


--
-----------------------------
Paulo Costa Carvalho, PhD
Laboratory for Structural and Computational Proteomics
Carlos Chagas Institute, Fiocruz - PR, Brazil
Journal of Proteomics - Executive Editor

Diogo Borges

unread,
Jun 8, 2019, 5:08:58 AM6/8/19
to si...@googlegroups.com
Hi Christian,

We've released a new version of SIM-XL (v. 1.5.3.5) that we've optimized the algorithm for large datasets.
In this case, the "Dynamic DB Reduction" option needs to be checked, and we recommend as XCorr threshold: 2.5 and Min. No. Peptides: 2.

This feature will perform a pre search (using Comet tool) to find out possible dead-end peptides. After that, a new database will be created with only proteins that contain at least one identified dead-end peptide.

If you have any questions, email us.

Cheers,
Diogo

PS: Even if this new feature reduces search time, the best alternative is to use cleavable cross-linkers, and we're already developing this new feature in our tool. As soon as possible we're gonna contact you to test it.

……………………………………………………………

Diogo Borges Lima | Institut Pasteur 

Post-doctoral fellow in Bioinformatics

Mass Spectrometry for Biology Unit - CNRS USR 2000

28 rue du Dr. Roux | 75724 Paris Cedex 15 | France

URL: diogobor | diogo.bo...@pasteur.fr

……………………………………………………………



To view this discussion on the web visit https://groups.google.com/d/msgid/simxl/CAOVP5VMdr6LCHqpDmM5Hj35WQZfo15ZuC7aXRtRMAUBubXW4yA%40mail.gmail.com.

TRAHAN Christian

unread,
Jun 10, 2019, 11:34:27 AM6/10/19
to si...@googlegroups.com

Thanks Diogo!

Aaron I.

unread,
Apr 30, 2021, 7:01:34 PM4/30/21
to SIM-XL Group
Hello,

Our group would also like to explore the use of SIM-XL in conjunction with cleavable crosslinkers for homo-oligomer complexes. Is the incorporation of cleavable crosslinkers still planned for SIM-XL?

Thank you.

Diogo Borges

unread,
Apr 30, 2021, 10:22:28 PM4/30/21
to SIM-XL
Dear Aaron,

Thanks for contacting us. Over the last months we have been working on the development of a new algorithm for identifying cleavable cross-linkers. This new algorithm has been massively tested and we hope it will be released soon.

With best regards,
Diogo
_____________________________________
Diogo Borges Lima
About me: @diogobor | diogobor
URL: diogobor

To view this discussion on the web visit https://groups.google.com/d/msgid/simxl/a44e32c1-4619-4baa-8020-bb6fedb4b3b5n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages