Open source for bioinformatics ()

Paul Shannon | 23 Apr 2012 18:35

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

(redirecting this back to the Bioc list…)

Hi Nooshin,

The 'bulk' approach is not quite so ready as I predicted.  I might have something by the end of the week.

As for mapping between PWMs and TFs, I have most often done this with 'tom-tom' from the meme website.  

But I just discovered what looks like a good -- maybe better -- approach:  the Bioconductor MotIV package,
which includes a 2010 version of jasper.
Try this:

   source("http://bioconductor.org/biocLite.R")

biocLite ('MotIV')
library (MotIV); 
browseVignettes ('MotIV')

The jaspar data in this package has 130 TF-PWM mappings, which appear to be human.  More must be known, and
publicly available.  The JASPAR website has a 'JASPAR CORE Plantae'  data set that 
   - is probably what you are interested in
   - might be downloadable, and convertible to the form MotIV wants.

Perhaps other readers of the list have other suggestions.

If you have any questions on this, please include 'BioC' in your reply, so that we can all get better at this!

 - Paul

On Apr 23, 2012, at 6:53 AM, nooshin wrote:

(Continue reading)

nooshin | 24 Apr 2012 10:01

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?


Hi Paul,

Thanks a lot.
I forgot to include bioc, since I only replied to you (no to all).

I can"t install MotIV package to check. I checked in google but I 
couldn't find any solution! Do you have any suggestion for installing 
this package?

Bests,
Nooshin

On 04/23/2012 06:35 PM, Paul Shannon wrote:
> (redirecting this back to the Bioc list)
>
> Hi Nooshin,
>
> The 'bulk' approach is not quite so ready as I predicted.  I might have something by the end of the week.
>
> As for mapping between PWMs and TFs, I have most often done this with 'tom-tom' from the meme website.
>
> But I just discovered what looks like a good -- maybe better -- approach:  the Bioconductor MotIV package,
which includes a 2010 version of jasper.
> Try this:
>
>     source("http://bioconductor.org/biocLite.R")
>
> biocLite ('MotIV')
> library (MotIV);

(Continue reading)

Tim Triche, Jr. | 24 Apr 2012 15:16

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

source("http://bioconductor.org/biocLite.R")
biocLite("MotIV")

ought to do the trick for you

On Tue, Apr 24, 2012 at 1:01 AM, nooshin <n_omranian@...> wrote:

>
> Hi Paul,
>
> Thanks a lot.
> I forgot to include bioc, since I only replied to you (no to all).
>
> I can"t install MotIV package to check. I checked in google but I
> couldn't find any solution! Do you have any suggestion for installing
> this package?
>
> Bests,
> Nooshin
>
> On 04/23/2012 06:35 PM, Paul Shannon wrote:
> > (redirecting this back to the Bioc list)
> >
> > Hi Nooshin,
> >
> > The 'bulk' approach is not quite so ready as I predicted.  I might have
> something by the end of the week.
> >
> > As for mapping between PWMs and TFs, I have most often done this with
> 'tom-tom' from the meme website.

(Continue reading)

nooshin | 24 Apr 2012 16:00

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Thanks, it's been already solved, it needs GSL package, which is a bit 
problematic, but I solved it already.

But it does include only 5 matrices (in the webpage) for arabidopsis and 
in the package also!
I'm downloading manually from AthaMap!

Thanks again and keep waiting for 'bulk' approach.

Bests,
Nooshin

On 04/24/2012 03:16 PM, Tim Triche, Jr. wrote:
> source("http://bioconductor.org/biocLite.R")
> biocLite("MotIV")
>
> ought to do the trick for you
>
>
>
> On Tue, Apr 24, 2012 at 1:01 AM, nooshin <n_omranian@... 
> <mailto:n_omranian@...>> wrote:
>
>
>     Hi Paul,
>
>     Thanks a lot.
>     I forgot to include bioc, since I only replied to you (no to all).
>

(Continue reading)

Tim Triche, Jr. | 24 Apr 2012 16:15

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Ah, I see.  GSL is a useful library to have installed regardless.  Hope
things work out.  I found your exchanges with Paul to be useful reading,
but obviously I was not reading closely enough, since Paul started off his
code sample with biocLite('MotIV').  Oops :-o

Here is a paper that I found interesting, which does go into some detail
towards a "bulk" approach, from Gottardo's group:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0016432

Perhaps it will be useful to you as well, would be curious to hear if so.

--t

On Tue, Apr 24, 2012 at 7:00 AM, nooshin <n_omranian@...> wrote:

> **
>
> Thanks, it's been already solved, it needs GSL package, which is a bit
> problematic, but I solved it already.
>
> But it does include only 5 matrices (in the webpage) for arabidopsis and
> in the package also!
> I'm downloading manually from AthaMap!
>
> Thanks again and keep waiting for 'bulk' approach.
>
> Bests,
> Nooshin
>

(Continue reading)

nooshin | 24 Apr 2012 16:36

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?


Thanks a lot for your suggestion. I will for sure have a look and inform 
you.
Bests,
Nooshin

On 04/24/2012 04:15 PM, Tim Triche, Jr. wrote:
> Ah, I see.  GSL is a useful library to have installed regardless. 
>  Hope things work out.  I found your exchanges with Paul to be useful 
> reading, but obviously I was not reading closely enough, since Paul 
> started off his code sample with biocLite('MotIV').  Oops :-o
>
> Here is a paper that I found interesting, which does go into some 
> detail towards a "bulk" approach, from Gottardo's group:
>
> http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0016432
>
> Perhaps it will be useful to you as well, would be curious to hear if so.
>
> --t
>
> On Tue, Apr 24, 2012 at 7:00 AM, nooshin <n_omranian@... 
> <mailto:n_omranian@...>> wrote:
>
>
>     Thanks, it's been already solved, it needs GSL package, which is a
>     bit problematic, but I solved it already.
>
>     But it does include only 5 matrices (in the webpage) for
>     arabidopsis and in the package also!

(Continue reading)

Mercier Eloi | 24 Apr 2012 19:28

Picon

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Hello,

I am one of the developer of MotIV. I will be happy to help you if you 
have any question regarding the package.

First, I want to mention that in the Plos One paper, we used PICS, 
rGADEM and MotIV as a pipeline but MotIV can be use as a stand alone. 
Some of the advanced functions won't be available though.

Since the PWMs in MotIV correspond to human TF, you may have to use your 
own list of PWMs. What MotIV needs is a simple list of matrices 
(head(jaspar) to view the format).
Jaspar's PWMs can be easily downloaded but it seems it only contains ~20 
motifs. On the other hand, AthaMap has more motifs but I did not manage 
to find an easy way to get them. Another place to look at is the AGRIS 
website (http://arabidopsis.med.ohio-state.edu/downloads.html).

If you're only interested by the identification of the motifs and do not 
want to do further analysis with R, I recommend you to look at 
http://www.benoslab.pitt.edu/stamp for the identification of your motifs.

Regards,

Eloi Mercier

On 12-04-24 07:36 AM, nooshin wrote:
> Thanks a lot for your suggestion. I will for sure have a look and inform
> you.
> Bests,
> Nooshin

(Continue reading)

Zhu, Lihua (Julie | 24 Apr 2012 19:47

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Eloi,

I would like to use MotIV for a c.elegans dataset. What data source would
you recommend for matchMotif? Many thanks for your help!

Best regards,

Julie

On 4/24/12 1:28 PM, "Mercier Eloi" <emercier@...> wrote:

> Hello,
> 
> I am one of the developer of MotIV. I will be happy to help you if you
> have any question regarding the package.
> 
> First, I want to mention that in the Plos One paper, we used PICS,
> rGADEM and MotIV as a pipeline but MotIV can be use as a stand alone.
> Some of the advanced functions won't be available though.
> 
> Since the PWMs in MotIV correspond to human TF, you may have to use your
> own list of PWMs. What MotIV needs is a simple list of matrices
> (head(jaspar) to view the format).
> Jaspar's PWMs can be easily downloaded but it seems it only contains ~20
> motifs. On the other hand, AthaMap has more motifs but I did not manage
> to find an easy way to get them. Another place to look at is the AGRIS
> website (http://arabidopsis.med.ohio-state.edu/downloads.html).
> 
> If you're only interested by the identification of the motifs and do not
> want to do further analysis with R, I recommend you to look at

(Continue reading)

Paul Shannon | 24 Apr 2012 21:02

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

The recent flurry of interest in sequence motifs here on the bioc list suggests to us that maybe we at
Bioconductor could strengthen our infrastructure for this kind of work.  If this work interests you --
either as a package creator, or as a package user -- please suggest ideas or use cases.  What do you need?  I
will collect and collate the responses.   We hope to identify places where Bioc can help out.

For background:  we already have a number of packages (rGADEM, MotIV, cosmo, BCRANK, motifRG) which
address, with different strengths, what I believe to be the two aspects of the motif problem:

  1) Detecting enriched motifs in DNA sequence, or in ChIP-seq data  (rGADEM, cosmo, motifRG, BCRANK)
  2) Predicting the sequence motifs which bind to these enriched motifs, and what binding molecules they
belong to (MotIV)

In the past, a lot of sequence motif/binding work has addressed the search for transcription factor
binding sites and their cognate transcription factors.  miRNAs, phorphorylation and methylation all
pose related problems.  Is there support which we can practically offer here as well?

In addition to Bioc packages, there are of course many worthwhile websites and external tools:  JASPAR,
meme, STAMP (and TRANSFAC, for those with a license).  Nooshin mentioned the arabidopsis-specific
'AthaMap' (http://www.athamap.de).  Are there other open-source data repositories like this for other
organisms?  c.elegans, as Julie requested?

Questions, suggestions, use cases and data sources are all welcome.

Thanks!

 - Paul

On Apr 24, 2012, at 10:47 AM, Zhu, Lihua (Julie) wrote:

> Eloi,

(Continue reading)

Zhu, Lihua (Julie | 25 Apr 2012 00:16

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Paul,

Thanks so much for the comprehensive summary of existing capability of Bioc
and other resources for motif discovery and matching!

Here is my response to your great initiative to collect use cases and open
data resources.

Here is an open data source for Drosophila which we developed:
http://pgfe.umassmed.edu/TFDBS/
http://nar.oxfordjournals.org/content/early/2010/11/19/nar.gkq858.full

As you pointed out, there are several excellent Bioconductor packages
available for the two common cases of motif problems, i.e., de nova motif
discovery and motif matching to known motifs. It would be useful to have
more motif databases available for motif comparison program such as MotIV.
In addition, we use clover to search for known motifs in a given set of
sequences.

Many thanks for sharing your insights!

Best regards,

Julie

On 4/24/12 3:02 PM, "Paul Shannon" <pshannon@...> wrote:

> The recent flurry of interest in sequence motifs here on the bioc list
> suggests to us that maybe we at Bioconductor could strengthen our
> infrastructure for this kind of work.  If this work interests you -- either as

(Continue reading)

Paul Shannon | 25 Apr 2012 05:02

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Hi Julie,

FlyFactorSurvey looks great.   Would that we had such a resource (curated, current, and growing) for all organisms!

A few questions, if I may:

  1) What role with respect to FlyFactorSurvey do you picture us taking here at BioC?  How can we help?

  2) Your website (http://pgfe.umassmed.edu/TFDBS) recommends meme and TOMTOM for motif comparison.  Do
you use them yourself?  If so, can you tell us about their strengths and weaknesses?  How do they compare to
clover?  (http://zlab.bu.edu/clover/)

In that same spirit -- trying to find out more about this topic -- here are some more questions:

  3) The JASPAR database seems to be mostly unchanged since 2009.  
     (http://jaspar.genereg.net/html/DOWNLOAD). Does anyone know their update policy? 

  4) Is TRANSFAC only for license holders?  

  5) Are there any other organism-specific gems like FlyFactorSurvey to be discovered out on the web?

Thanks!

 - Paul  

On Apr 24, 2012, at 3:16 PM, Zhu, Lihua (Julie) wrote:

> Paul,
> 
> Thanks so much for the comprehensive summary of existing capability of Bioc

(Continue reading)

nooshin | 25 Apr 2012 10:25

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Hi All,

I followed all your emails, Many thanks for the nice information you shared.

3) The JASPAR database seems to be mostly unchanged since 2009.
      (http://jaspar.genereg.net/html/DOWNLOAD). Does anyone know their update policy?

No, I don't know, I just checked for arabidobsis thaliana and could find 
only FIVE genes!

4) Is TRANSFAC only for license holders?

Yes, one should pay for it in order to be able to use!

   5) Are there any other organism-specific gems like FlyFactorSurvey to be discovered out on the web?

This is motif sampler, which I found before, I'm not sure if this works 
like FlyFactorSurvey or not
http://homes.esat.kuleuven.be/~sistawww/bioi/thijs/Work/MotifSampler.html

Many thanks Paul for trying to clarify all these points!
Bests,
Nooshin

On 04/25/2012 05:02 AM, Paul Shannon wrote:
> Hi Julie,
>
> FlyFactorSurvey looks great.   Would that we had such a resource (curated, current, and growing) for all organisms!
>
> A few questions, if I may:

(Continue reading)

Zhu, Lihua (Julie | 25 Apr 2012 15:44

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Paul,

Thanks for the positive feedback on FlyFactorSurvey! The motifs in this
database are generated using the bacterial one-hybrid method (B1H and
B1H-seq). All the public motifs can be downloaded freely. It would be useful
to have a Bioc data package, containing curated and current motifs from all
organisms if available, that interfaces with MotiV.

MEME works very well in finding motifs from B1H-seq data (Christensen et
al.,Nucleic Acid Research 2011, Vol39, No.12 e83), although only limited
motif discovery tools were compared in the paper. Currently, we are working
on whether motif discovery can be improved with B1H-seq data.

As I understand, MEME is for de nova motif discovery, TOMTOM and STAMP are
for testing whether the motif returned by a motif finder is significantly
similar to a known motif, clover is for searching known motifs in a given
set of sequences. We are thinking of adding clover to our website.

I am looking forward to your collated survey results.

Best regards,

Julie

On 4/24/12 11:02 PM, "Paul Shannon" <pshannon@...> wrote:

> Hi Julie,
> 
> FlyFactorSurvey looks great.   Would that we had such a resource (curated,
> current, and growing) for all organisms!

(Continue reading)

Steve Lianoglou | 25 Apr 2012 16:12

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Hi,

To carry on the MEME stuff, a biostar post just pointed me to an
updated scoring metric in tomtom which is made available in the latest
MEME software suite:

http://bioinformatics.oxfordjournals.org/content/27/12/1603.full

Perhaps wrapping parts of the MEME suite into an R library would be useful, no?

You might find the FIRE (and FIRE-pro) suite of tools also useful for
motif discovery, as welll:

http://physiology.med.cornell.edu/faculty/elemento/lab/software.shtml

Related to that, S. Tavazoie gave a talk at the recent CSHL/sysbio
meeting and presented TEISER, which seems pretty cool if you're
looking for structural motifs:

https://tavazoielab.c2b2.columbia.edu/TEISER/

-steve

On Wed, Apr 25, 2012 at 9:44 AM, Zhu, Lihua (Julie)
<Julie.Zhu@...> wrote:
> Paul,
>
> Thanks for the positive feedback on FlyFactorSurvey! The motifs in this
> database are generated using the bacterial one-hybrid method (B1H and
> B1H-seq). All the public motifs can be downloaded freely. It would be useful

(Continue reading)

Pj Dias | 25 Apr 2012 20:25

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Regarding S. cerevisiae, there is a great resource, YEASTRACT database (
http://www.yeastract.com/).

It compiles the motifs already identified, documented regulation based on
wet-lab research and potential regulation based on comparison of motifs
against the promoter region of genes.

It is reviewed and up-dated regularly, being added new motifs and new
documented regulation.

pj

No dia 25 de Abril de 2012 15:12, Steve Lianoglou <
mailinglist.honeypot@...> escreveu:

> Hi,
>
> To carry on the MEME stuff, a biostar post just pointed me to an
> updated scoring metric in tomtom which is made available in the latest
> MEME software suite:
>
> http://bioinformatics.oxfordjournals.org/content/27/12/1603.full
>
> Perhaps wrapping parts of the MEME suite into an R library would be
> useful, no?
>
> You might find the FIRE (and FIRE-pro) suite of tools also useful for
> motif discovery, as welll:
>
> http://physiology.med.cornell.edu/faculty/elemento/lab/software.shtml

(Continue reading)

Arno BioC | 27 Apr 2012 14:00

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?

Hi,

We are the developers of the de novo motif discovery package rGADEM. While rGADEM fits nicely in the
ChIP-Seq analysis pipeline PICS/rGADEM/MotIV, it can also easily be used to analyze enriched regions
obtained by any other peak callers. Among the many output of rGADEM, we find the position weight matrices
(pwn) that can be directly used for analysis in MotIV, but also other programs that compare motif
sequences with databases. Since it is implemented using OpenMP, the time for each analysis required for
every analysis is greatly reduced.

We would happily consider collaborating with you if you are interested to test our de novo motif discovery
package with your dataset.

Regards,

Arnaud.

Le 2012-04-25 à 10:12, Steve Lianoglou a écrit :

> Hi,
> 
> To carry on the MEME stuff, a biostar post just pointed me to an
> updated scoring metric in tomtom which is made available in the latest
> MEME software suite:
> 
> http://bioinformatics.oxfordjournals.org/content/27/12/1603.full
> 
> Perhaps wrapping parts of the MEME suite into an R library would be useful, no?
> 
> You might find the FIRE (and FIRE-pro) suite of tools also useful for
> motif discovery, as welll:

(Continue reading)

nooshin | 25 Apr 2012 10:08

Picon

Re: Motif search -- access to JASPAR, MotIV package, more TF-PWM relationships?


Hello,

Many thanks for the reply.
Actually, I'm working with Arabidopsis thaliana, and this is my first 
time to do some sequence-based analysis. I have a list of TFs, and some 
clusters of genes, I need to check which TFs' binding site(motifs) are 
enriched in these clusters. The other analysis is that, I have a TF in 
Arabidopsis which includes two part (RAV1) and I need to check if the 
motif of this specific TF is enriched in any of the genes' sequences 
(500~1000-bps) in these clusters.
I need to know which packages in R is helpful to use? since it's really 
my first experience of the sequence-based analysis.
I took all information in AthaMap manually and restructured them, both 
matrix-based and pattern-based. I sent them an email and they've not 
supplied me with any data yet, so I did it manually!
I also checked AGRIS, but I couldn't find any matrix-based information 
there :(
I used stamp, but with the STAMP there is no possibility to give a 
specific motif and search for its enrichment in sequences!
Also, as you mentioned, I need to do further analysis with R after 
identification of the motifs, so I prefer to use R instead of using 
online tools.
I also checked MEME (MAST and MEME-ChIP), which I could get the result 
which I need, but I need to work with R.

Many thanks all.
Bests,
Nooshin

(Continue reading)

Return

Return to gmane.science.biology.informatics.conductor.

Project Web Page

Open source for bioinformatics ()

Search Archive

Language

Change language

Options

Current view: Threads only / Showing only 20 lines / Not hiding cited text.
Change to All messages, whole messages, or hide cited text.

Post a message
NNTP Newsgroup
Classic Gmane web interface
XML

XML

RSS Feed
List Information

About Gmane