genex-dev — GeneX developers

[GeneX-dev] Question about Using Curation Tool

[Shih-hsiu L. <ma...@jh...>]

To whom it may concert,
	On the Curation Tool, I clicked "CREATE NEW EXPERIMENT SET", then
clicked on "DOWNLOAD CONTROL FILES". Then the brower opened a web page
that required login name and password. I wonder how I can register an
account. Thank you.


Sincerely,

Shih-Hsiu Lin

Re: [GeneX-dev] GeneX Installation Problem

[<ja...@op...>]

"Shih-hsiu Lin" <ma...@jh...> writes:

> 	I have attached a file, logfile, containing the error messages
> when I ran the statment: zcat genex-initdb-latest.gz |psql -U genex.
> It seems like I have problem access the database. Please give me some
> sugessions. Thank you.

Hi Shih-hsiu,

You must execute the psql command with a user who has permission to
modify your Postgres installation since it's going to create tables. 

Also, that script should be run for you by the installer, and it will
create all the users that need to exist, as well as setting the DB
permissions for you. Why are you running it by hand? Did you have
problems running install-all.pl?

jas.

[GeneX-dev] GeneX Installation Problem

[Shih-hsiu L. <ma...@jh...>]

To whom it may concert,
	I have attached a file, logfile, containing the error messages
when I ran the statment: zcat genex-initdb-latest.gz |psql -U genex.
It seems like I have problem access the database. Please give me some
sugessions. Thank you.


Sincerely,

Shih-Hsiu Lin

[GeneX-dev] Re: GeneX Installation Problems

[Todd P. <tf...@nc...>]

Note that I have cc'd the GeneX sourceforge list, please respond to that
address since it reaches a broader base of GeneX experts.

We need to know if it is an operating-system (O/S) error message or if it is
a database error message. I doubt the first case, since the command isn't
writing a file. So, it is probably the second case (DB access error). you
must be able to use the psql command by itself by typing
  psql -U genex
(looks like the -U is missing from your command). The Postgres SQL console
should come up without any problems. The Postgres database needs a user
called genex which is O/S verified by looking at your user group (type
'groups' to see which groups you belong to). If you belong the the genex O/S
group, psql -U genex will start up the Postgres SQL console without
prompting for a password. If you log on to the O/S as genex, you do not need
to supply the -U genex parameters to psql. Once that works right try
zcat genex-initdb-latest.gz | psql -U genex

Let us know if that doesn't work.

Todd
----- Original Message -----
From: "Shih-hsiu Lin" <ma...@jh...>
To: "Todd F. Peterson" <tf...@nc...>
Sent: Tuesday, August 07, 2001 9:05 AM
Subject: GeneX Installation Problems


> To whom it may concert,
> When I ran the following statement:
> zcat genex-initdb-latest.gz |psql genex
> it caused many Permission Denied errors. I have changed the owner
> and group of /usr/local/pgsql/ directory to postgres. It didn't solve the
> problem. Please give me some suggesions. Thank you.
>
>
> Sincerely,
>
> Shih-Hsiu Lin
>
>
>

Re: [GeneX-dev] Re: data import?

[<ja...@op...>]

Todd's answer is corrct, if such a GATC -> GeneXML tool existed. As
far as I know, there is no such tool. So far all practical purposes
the answer is "No". 

However, there is hope:

The Object Management Group (OMG) has a working group on gene
expression data, of which Rosetta, GeneLogic, and MGED are all
submittors. We will have a revised draft of the MicroArray Gene
Exression Object Model (MAGE-OM) and the MicroArray Gene
Exression Markup Language (MAGE-ML) by the 20th of Aug. 

This will replace Rosetta's GEML and MGED's MAML as standard
transmission formats. Affymetrix has been a participant in this
process and knows all about it. As soon as it becomes a standard OMG
specification they will support it. 

There will be a programming jamboree in Toronto Sept 14-19, in which
individuals from academia and industry have agreed to participate to
write Open Source tools which can import/export MAGE-ML, as well as
design relational schema's to support MAGE-OM, and writing Java and
Perl API's for such databases.

If *you* want to see GeneX support GATC much sooner, get your company
to sponser a single programmer for the jamboree, and have the person
help write a GATC --> MAGE-ML translator. This will benefit your
company and the whole world.

jas. 

"Todd F. Peterson" <tf...@nc...> writes:

> Yes, with a tool that can translate the relevant data from GATC to the
> proper XML format which GeneX understands. I have cc'd the sourceforge
> list on this message which reaches a broader audience of GeneX experts.
> 
> Todd Peterson-Sofware Developer
> NCGR
> 
> On Fri, 3 Aug 2001 Qi....@as... wrote:
> 
> > Hi,
> >
> > Can GeneX import data directly from GATC database?
> >
> > Qi Zhang
> > EST informatics
> > AstraZeneca

[GeneX-dev] Re: data import?

[Todd F. P. <tf...@nc...>]

mailed to wrong address first time.

todd

Date: Fri, 3 Aug 2001 16:32:59 -0600 (MDT)
From: Todd F. Peterson <tf...@nc...>
To: Qi....@as...
Cc: ge...@li..., ge...@nc...
Subject: Re: data import?

Yes, with a tool that can translate the relevant data from GATC to the
proper XML format which GeneX understands. I have cc'd the sourceforge
list on this message which reaches a broader audience of GeneX experts.

Todd Peterson-Sofware Developer
NCGR

On Fri, 3 Aug 2001 Qi....@as... wrote:

> Hi,
>
> Can GeneX import data directly from GATC database?
>
> Qi Zhang
> EST informatics
> AstraZeneca
>

Re: [GeneX-dev] query problem

[<ja...@op...>]

"Ron Ophir" <Ls...@wi...> writes:

> /([-\w!\#$%&'*+.^_\`|{}~]+):\s+([^]*)/: unmatched [] in regexp at

I found the problem. This is from CGI.pm in the read_header
subroutine. What version are you using (I've got 2.70)? To figure that
out use:

  /path/to/your/perl -MCGI -e 'print "$CGI::VERSION\n"'

You either need to upgrade or downgrade your CGI.pm. Let me know the
version, and I'll send a bug report to Lincoln Stein.

Thanks Ron,
jas.

Re: [GeneX-dev] query problem

[<ja...@op...>]

"Ron Ophir" <Ls...@wi...> writes:

> I re-install genex on a new hardware that I got. When I query for
> one of the experiment I got this:
> /([-\w!\#$%&'*+.^_\`|{}~]+):\s+([^]*)/: unmatched [] in regexp at
> (eval 8) line 100.  GeneX DB Query 2/4: Experiment Selection Nothing
> to filter or select on - please go back and choose at least one
> selection!

Hey Ron,

The error message is telling you there is a bad regular expression in
that line. It looks legal to me, and it certainly works on my machine
and others. I tried to login into your system to check your
configuration, but you have it nicely password protected, so I
couldn't. What version of perl are you running (use perl -v).

thanks,
jas.

> running
> http://wisard.weizmann.ac.il/cgi-bin/genex/gxquery/fetch_es.pl
> What is the reason?
> 
> thanks,
> Ron

Re: [GeneX-dev] query problem

[Todd P. <tf...@nc...>]

Could you provide the specific sequence of operations you used to get =
this message? Are you using your own experiment set or the sample one's =
already in the database? Did you go back and choose one selection as the =
message requests?

Todd Peterson-Software Developer
  ----- Original Message -----=20
  From: Ron Ophir=20
  To: gen...@li...=20
  Sent: Wednesday, August 01, 2001 7:34 AM
  Subject: [GeneX-dev] query problem


  hi all,
  I re-install genex on a new hardware that I got. When I query for one =
of the experiment I got this:
  /([-\w!\#$%&'*+.^_\`|{}~]+):\s+([^]*)/: unmatched [] in regexp at =
(eval 8) line 100.
  GeneX DB Query 2/4: Experiment Selection
  Nothing to filter or select on - please go back and choose at least =
one selection!=20

  running
  http://wisard.weizmann.ac.il/cgi-bin/genex/gxquery/fetch_es.pl
  What is the reason?

  thanks,
  Ron

[GeneX-dev] query problem

[Ron O. <Ls...@wi...>]

hi all,
I re-install genex on a new hardware that I got. When I query for one of =
the experiment I got this:
/([-\w!\#$%&'*+.^_\`|{}~]+):\s+([^]*)/: unmatched [] in regexp at (eval 8) =
line 100.
GeneX DB Query 2/4: Experiment Selection
Nothing to filter or select on - please go back and choose at least one =
selection!=20

running
http://wisard.weizmann.ac.il/cgi-bin/genex/gxquery/fetch_es.pl
What is the reason?

thanks,
Ron

[GeneX-dev] Re: GeneX Installation Problems

[Todd P. <tf...@nc...>]

Dear Shih-hsiu Lin:

Did you create and populate the Postgres database before you ran =
install-all.pl (see the prerequisite requirements in the file called =
INSTALL which is at the root of the installation file tree.)? The error =
you have reported looks like the database is not installed and/or =
populated correctly. Also, be sure you have the most recent version =
located at http://genex.ncgr.org/genex/download/genex-server/.=20

Please let us know if you have any other problems or questions. Reply to =
gen...@li... since that address reaches a broader =
audience of GeneX experts.

Thank you,
Todd Peterson
Software Developer
tf...@nc...



Date: Mon, 30 Jul 2001 13:46:39 -0400
From: Shih-hsiu Lin <ma...@jh...>
Subject: Question about GeneX
To: ge...@nc...
MIME-version: 1.0

To whom it may concert,
I got an error message when I tried to run install-all.pl. The
following is the error message:

DBI->connect(dbname=3Dtemplate1) failed: connectDB() -- connect()
failed: Connecti on refused at install-all.pl line 1574
Can't call method "prepare" on an undefined value at install-all.pl line
1578, <STDIN> chunk 22.

Please inform me what I should do to fix the problem. Thank you.


Sincerely,

Matthew Lin

Re: [GeneX-dev] Re: GeneX installation

[Bill B. <wd...@nc...>]

Harry,

I've been watching the interchange, hopefully you have soothed his ego;
Based on his AMAD experience he does seem like a good potential contributor
to GeneX.  email is a terrible way to communicate, but we are stuck with it.
Hang in there.

-wdb

----- Original Message -----
From: "Harry Mangalam" <man...@ho...>
To: "Chao Lu" <ch...@uh...>
Cc: "GeneX" <gen...@li...>
Sent: Friday, July 13, 2001 8:06 AM
Subject: [GeneX-dev] Re: GeneX installation


> Hi Chao,
>
> Chao Lu wrote:
> >
> > Dear Harry:
> >
> > My initial request of help is sent to ge...@nc..., which is listed in
> > some of your document.
> > I did not mean to to bother particular person, like you to answer my
questions.
>
> OK - that's fair; it's just that many of us have other commitments and if
one person on the list can't answer, others will.
>
> > Regarding the installation, apparently, I have to spend more time
reading
> > it to make sure that I
> > understand what you mean in the doc. However, I do not think that you
have
> > a good instructions.
>
> We are always looking for suggestions (and editors..)  Most of the
suggestions that we have gotten have been included in the docs.
>
> > What I copied to you is the part that I think is important (BEFORE
> > CONFIGURE GENEX), and I
> > think that I should run the first phase first, not the second phase
first.
> > Instead, you asked me to run the second phase first.
>
> I'm sorry - that was because I assumed you /had/ run the 1st part already;
the installation script checks for the 1st phase being complete and tries to
do it for you as well if it was not done, but it
> can fail and so it's smoother if you do it in order.
>
>
> > I have to admit that I did not read the whole doc of Readme, since you
have
> > a doc of Install. I thought
> > everything on installation should be in the install file.
>
> Everything for the installation IS in the INSTALL file, but the README
gives an overview that helps to understand the system better.
>
> > I know you and your colleagues have used too many of PLEASE, READ, THE
> > WHOLE, BOTH in the
> > limited communication with me. Instead, why don't you get some one to
make
> > the document easier
> > to read.
>
> Forgive the shouting, but I did this as it was clear that you /weren't/
reading the instructions.  If you would like to help out by noting how this
would be done, we'd be happy for your assistance -
> this is an Open Source project and we're stretched a bit thin.  I do admit
that documentation often seems to take a second seat to code and
functionality, but we're trying..
>
> > PLEASE DO NOT spread installation instruction into TWO documents. This
is
> > my suggestion.
>
> OK - but the breaking of doc into two files is an accepted practice with
most software packages.
>
> > Apparently, GeneX is free. I should not expect any more than just
"READ". I
> > will READ everything to
> > see if I can get it works.
>
> On the contrary, please come back to us with questions and comments.  But
we all have day jobs writing the underlying code, so if we've already
answered your questions in a relatively straightforward
> way in the documentation, we feel that you should do us the courtesy of
reading it at least once.
>
> As we make clear several times, GeneX is NOT a trivial system to install.
Unlike AMAD which is a flat file database, GeneX is based on a relational
database.  This makes it more difficult to install,
> but also more powerful in the end.
>
> > By the way, I am not a person that cannot understand programming terms
to
> > understand your docs. I spent no
> > time to install the AMAD into my system. Since then, I modified almost
all
> > the Perl scripts for our microarrays
> > and I also added many functions that AMAD did not have.
>
> that's also great - I noted that you probably WERE a knowledgable person
with computers - just the kind of person we'd like to have interested in
GeneX.
>
> AMAD can install this way as it it is strictly CGIs and HTML/Javascript,
so there are only 2 place to install things.  Not to say anything against
AMAD and we like and interact with its developers -
> it certainly has a nicer interface than GeneX, but the two are different
animals.
>
> --
> Cheers,
> Harry
>
> Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...
>
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev
>

[GeneX-dev] Re: GeneX installation

[Harry M. <man...@ho...>]

Hi Chao,

Chao Lu wrote:
> 
> Dear Harry:
> 
> My initial request of help is sent to ge...@nc..., which is listed in
> some of your document.
> I did not mean to to bother particular person, like you to answer my questions.

OK - that's fair; it's just that many of us have other commitments and if one person on the list can't answer, others will.

> Regarding the installation, apparently, I have to spend more time reading
> it to make sure that I
> understand what you mean in the doc. However, I do not think that you have
> a good instructions.

We are always looking for suggestions (and editors..)  Most of the suggestions that we have gotten have been included in the docs.

> What I copied to you is the part that I think is important (BEFORE
> CONFIGURE GENEX), and I
> think that I should run the first phase first, not the second phase first.
> Instead, you asked me to run the second phase first.

I'm sorry - that was because I assumed you /had/ run the 1st part already; the installation script checks for the 1st phase being complete and tries to do it for you as well if it was not done, but it
can fail and so it's smoother if you do it in order.


> I have to admit that I did not read the whole doc of Readme, since you have
> a doc of Install. I thought
> everything on installation should be in the install file.

Everything for the installation IS in the INSTALL file, but the README gives an overview that helps to understand the system better.

> I know you and your colleagues have used too many of PLEASE, READ, THE
> WHOLE, BOTH in the
> limited communication with me. Instead, why don't you get some one to make
> the document easier
> to read.

Forgive the shouting, but I did this as it was clear that you /weren't/ reading the instructions.  If you would like to help out by noting how this would be done, we'd be happy for your assistance -
this is an Open Source project and we're stretched a bit thin.  I do admit that documentation often seems to take a second seat to code and functionality, but we're trying..

> PLEASE DO NOT spread installation instruction into TWO documents. This is
> my suggestion.

OK - but the breaking of doc into two files is an accepted practice with most software packages.

> Apparently, GeneX is free. I should not expect any more than just "READ". I
> will READ everything to
> see if I can get it works.

On the contrary, please come back to us with questions and comments.  But we all have day jobs writing the underlying code, so if we've already answered your questions in a relatively straightforward
way in the documentation, we feel that you should do us the courtesy of reading it at least once. 

As we make clear several times, GeneX is NOT a trivial system to install.  Unlike AMAD which is a flat file database, GeneX is based on a relational database.  This makes it more difficult to install,
but also more powerful in the end.

> By the way, I am not a person that cannot understand programming terms to
> understand your docs. I spent no
> time to install the AMAD into my system. Since then, I modified almost all
> the Perl scripts for our microarrays
> and I also added many functions that AMAD did not have.

that's also great - I noted that you probably WERE a knowledgable person with computers - just the kind of person we'd like to have interested in GeneX.  

AMAD can install this way as it it is strictly CGIs and HTML/Javascript, so there are only 2 place to install things.  Not to say anything against AMAD and we like and interact with its developers -
it certainly has a nicer interface than GeneX, but the two are different animals.

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...

[GeneX-dev] Re: GeneX installation

[Chao Lu <ch...@uh...>]

Dear Harry:

My initial request of help is sent to ge...@nc..., which is listed in 
some of your document.
I did not mean to to bother particular person, like you to answer my questions.

Regarding the installation, apparently, I have to spend more time reading 
it to make sure that I
understand what you mean in the doc. However, I do not think that you have 
a good instructions.

What I copied to you is the part that I think is important (BEFORE 
CONFIGURE GENEX), and I
think that I should run the first phase first, not the second phase first.
Instead, you asked me to run the second phase first.

I have to admit that I did not read the whole doc of Readme, since you have 
a doc of Install. I thought
everything on installation should be in the install file.

I know you and your colleagues have used too many of PLEASE, READ, THE 
WHOLE, BOTH in the
limited communication with me. Instead, why don't you get some one to make 
the document easier
to read.

PLEASE DO NOT spread installation instruction into TWO documents. This is 
my suggestion.

Apparently, GeneX is free. I should not expect any more than just "READ". I 
will READ everything to
see if I can get it works.

By the way, I am not a person that cannot understand programming terms to 
understand your docs. I spent no
time to install the AMAD into my system. Since then, I modified almost all 
the Perl scripts for our microarrays
and I also added many functions that AMAD did not have.

Thanks.



Chao Lu, M.D.,Ph.D.
Coordinator,
Microarray Lab at Ontario Cancer Institute

=======================================================


At 06:28 PM 12/07/2001 -0700, you wrote:
>The part you copied in is not the entire document.  If you're not willing 
>to read the WHOLE of BOTH the README and the INSTALL documents, then you 
>probably shouldn't try to install this package.  It
>is relatively straightforward, but it is NOT trivial.
>
>Please READ the docs.  If you don't understand large parts, then you 
>shouldn't install it, but if you installed and modified AMAD, then you 
>should have no trouble with this.
>
>GeneX requires a running PostgreSQL database and to have a database 
>instantiated with the script referreed to below 'init-dump.pl'.  If you 
>can't get that up and running, there's little point in
>continuing.
>
>And PLEASE post request to <gen...@li...>.  You'll get 
>wider readership and help than if you post only to me.
>
>Best wishes
>Harry
>
>
>
>
>Chao Lu wrote:
> >
> > Hi, Harry:
> >
> > Here is your doc. What pl should run first ?
> >
> >
> > privilege: The two phases of installation require different user privileges
> >
> >   * The first phase, installing the Pg DB dump (init-dump.pl) must be
> >     run as a user with permission to create new databases and users
> >     within your Postgres installation.  This is usually the user 
> 'postgres', if
> >     the DB was part of a RedHat or Debian installation.
> >
> >     N O T E:  When you untar the server package, the directory itself 
> has to be
> >     EXECUTABLE by ALL, or the 'init-dump.pl' script will fail. If it is not
> >     mode 755, make it so with the command:
> >        chmod 755 GeneX-Server-X.Y.Z
> >
> >   * The second phase general configuration and installation (using the
> >     'install-all.pl' script) still needs to be run by a privileged user
> >     (usually root) who can write to system directories (the root of 
> your CGI
> >     tree, and the root of your html document tree).
> >
> > At 12:16 PM 12/07/2001 -0700, you wrote:
> >
> > > PLEASE read the INSTALLATION documentation.  You have to run the 
> script 'install-all.pl' 1st.
> > >
> > > hjm
> > >
> > > Chao Lu wrote:
> > > >
> > > > Hi, Harry:
> > > >
> > > > Thank you for your quick reply.
> > > > Here is the error message I got when I typed init-dump.pl.in to run 
> the script.
> > > >
> > > > soleil[38] init-dump.pl.in
> > > > init-dump.pl.in: %%START_PERL%%: not found
> > > > init-dump.pl.in: use: not found
> > > > init-dump.pl.in: syntax error at line 3: `(' unexpected
> > > >
> > > > Chao
> > >
> > > --
> > > Cheers,
> > > Harry
> > >
> > > Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...
>
>--
>Cheers,
>Harry
>
>Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...

[GeneX-dev] Re: GeneX installation

[Harry M. <man...@ho...>]

The part you copied in is not the entire document.  If you're not willing to read the WHOLE of BOTH the README and the INSTALL documents, then you probably shouldn't try to install this package.  It
is relatively straightforward, but it is NOT trivial.

Please READ the docs.  If you don't understand large parts, then you shouldn't install it, but if you installed and modified AMAD, then you should have no trouble with this.

GeneX requires a running PostgreSQL database and to have a database instantiated with the script referreed to below 'init-dump.pl'.  If you can't get that up and running, there's little point in
continuing.

And PLEASE post request to <gen...@li...>.  You'll get wider readership and help than if you post only to me.

Best wishes
Harry



 
Chao Lu wrote:
> 
> Hi, Harry:
> 
> Here is your doc. What pl should run first ?
> 
> 
> privilege: The two phases of installation require different user privileges
> 
>   * The first phase, installing the Pg DB dump (init-dump.pl) must be
>     run as a user with permission to create new databases and users
>     within your Postgres installation.  This is usually the user 'postgres', if
>     the DB was part of a RedHat or Debian installation.
> 
>     N O T E:  When you untar the server package, the directory itself has to be
>     EXECUTABLE by ALL, or the 'init-dump.pl' script will fail. If it is not
>     mode 755, make it so with the command:
>        chmod 755 GeneX-Server-X.Y.Z
> 
>   * The second phase general configuration and installation (using the
>     'install-all.pl' script) still needs to be run by a privileged user
>     (usually root) who can write to system directories (the root of your CGI
>     tree, and the root of your html document tree).
> 
> At 12:16 PM 12/07/2001 -0700, you wrote:
> 
> > PLEASE read the INSTALLATION documentation.  You have to run the script 'install-all.pl' 1st.
> >
> > hjm
> >
> > Chao Lu wrote:
> > >
> > > Hi, Harry:
> > >
> > > Thank you for your quick reply.
> > > Here is the error message I got when I typed init-dump.pl.in to run the script.
> > >
> > > soleil[38] init-dump.pl.in
> > > init-dump.pl.in: %%START_PERL%%: not found
> > > init-dump.pl.in: use: not found
> > > init-dump.pl.in: syntax error at line 3: `(' unexpected
> > >
> > > Chao
> >
> > --
> > Cheers,
> > Harry
> >
> > Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...

Re: [GeneX-dev] Re: GeneX installation

[<ja...@op...>]

Hello Chao,

There is a document in the distribution entitled, 'INSTALL' and one
entitled 'README'. Please read them both. The install procedure is
complicated enough without trying to do it by braille.

Also please reply to the list, gen...@li..., and
not directly to Harry. You're are more likely to get help if you post
so that all the developers can read it.

jas.


"Harry Mangalam" <man...@ho...> writes:

> PLEASE read the INSTALLATION documentation.  You have to run the script 'install-all.pl' 1st.
> 
> hjm
> 
> Chao Lu wrote:
> > 
> > Hi, Harry:
> > 
> > Thank you for your quick reply.
> > Here is the error message I got when I typed init-dump.pl.in to run the script.
> > 
> > soleil[38] init-dump.pl.in
> > init-dump.pl.in: %%START_PERL%%: not found
> > init-dump.pl.in: use: not found
> > init-dump.pl.in: syntax error at line 3: `(' unexpected
> > 
> > Chao
> 
> -- 
> Cheers,
> Harry
> 
> Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...
> 
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev

[GeneX-dev] Re: GeneX installation

[Harry M. <man...@ho...>]

PLEASE read the INSTALLATION documentation.  You have to run the script 'install-all.pl' 1st.

hjm

Chao Lu wrote:
> 
> Hi, Harry:
> 
> Thank you for your quick reply.
> Here is the error message I got when I typed init-dump.pl.in to run the script.
> 
> soleil[38] init-dump.pl.in
> init-dump.pl.in: %%START_PERL%%: not found
> init-dump.pl.in: use: not found
> init-dump.pl.in: syntax error at line 3: `(' unexpected
> 
> Chao

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...

Re: [GeneX-dev] Re: GeneX installation

[Bill B. <wd...@nc...>]

Harry,

You provide great support for this project.  Thanks, Bill

----- Original Message -----
From: "Harry Mangalam" <man...@ho...>
To: "Chao Lu" <ch...@uh...>
Cc: "genexdev at SF" <gen...@li...>
Sent: Thursday, July 12, 2001 9:46 AM
Subject: [GeneX-dev] Re: GeneX installation


> Hi there,
>
> You need to read more about the installation procedure - DO NOT unpack and
execute this installation in your cgi-bin (or it'll be flooded with all
kinds of files and directories).
>
> Please unpack the installation tarball in a separate directory and then
see the README and INSTALL files that came with it.  If you have questions
about it after that, we'll be happy to help you out.
>
>
>
> Chao Lu wrote:
> >
> > To whom it may concern:
> >
> > I just found your GeneX project and downloaded the files to my cgi-bin
dir
> > since I thought your install-all.pl
> > may need cgi-bin.
> >
> > The question I have before I run your install-all.pl  is that you design
> > GeneX in the way that scripts
> > are separated from docs (such as html, txt, xml and data tables) or they
> > are under the same dir ?
>
> CGI scripts, HTML docs, are located in your web directory as normal and
are put there in response to the locations you specify in the installation
procedures.  Other libraries, docs, etc are placed
> together in another genex directory.
>
> > I took a quick look at your Genex dir, sub dir contains a lot pm or pl.
> > Should they be in cgi-bin, or html doc
> > dir?
>
> Please read the docs - you have to run the install-all.pl script to
install everything or you'll be living in a world of pain for a long time.
;)
>
> > Recently, I modified a lot on Stanford's AMAD scripts for my project and
> > new version, OCI AMAD
> > (what I call, OCI''s version of AMAD) works fine for our microarrays.
> > http://www.uhnres.utoronto.ca/services/microarray/amad/index.html
> >
> > Our microarray centre is producing almost the most of microarrays to
> > academic users world wide.
>
> You're using AMAD to host more data than SMD (>15000 experiments)?!?!
Congratulations!
>
> > I think that there is a need for me to get to know your GeneX system to
see
> > if I can modify it to
> > our microarrays.
>
> We'd be happy to help you, but please read the installation docs.
>
> > Sincerely,
> >
> > Chao Lu
> > Microarray Centre at Ontario Cancer Institute.
>
> --
> Cheers,
> Harry
>
> Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...
>
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev
>

[GeneX-dev] Re: GeneX installation

[Harry M. <man...@ho...>]

Hi there,

You need to read more about the installation procedure - DO NOT unpack and execute this installation in your cgi-bin (or it'll be flooded with all kinds of files and directories).

Please unpack the installation tarball in a separate directory and then see the README and INSTALL files that came with it.  If you have questions about it after that, we'll be happy to help you out.



Chao Lu wrote:
> 
> To whom it may concern:
> 
> I just found your GeneX project and downloaded the files to my cgi-bin dir
> since I thought your install-all.pl
> may need cgi-bin.
> 
> The question I have before I run your install-all.pl  is that you design
> GeneX in the way that scripts
> are separated from docs (such as html, txt, xml and data tables) or they
> are under the same dir ?

CGI scripts, HTML docs, are located in your web directory as normal and are put there in response to the locations you specify in the installation procedures.  Other libraries, docs, etc are placed
together in another genex directory.

> I took a quick look at your Genex dir, sub dir contains a lot pm or pl.
> Should they be in cgi-bin, or html doc
> dir?

Please read the docs - you have to run the install-all.pl script to install everything or you'll be living in a world of pain for a long time. ;)

> Recently, I modified a lot on Stanford's AMAD scripts for my project and
> new version, OCI AMAD
> (what I call, OCI''s version of AMAD) works fine for our microarrays.
> http://www.uhnres.utoronto.ca/services/microarray/amad/index.html
> 
> Our microarray centre is producing almost the most of microarrays to
> academic users world wide.

You're using AMAD to host more data than SMD (>15000 experiments)?!?! Congratulations!

> I think that there is a need for me to get to know your GeneX system to see
> if I can modify it to
> our microarrays.

We'd be happy to help you, but please read the installation docs.

> Sincerely,
> 
> Chao Lu
> Microarray Centre at Ontario Cancer Institute.

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- man...@ho...

[GeneX-dev] Re: CyberT

[Harry M. <man...@ho...>]

Rachel Tompa wrote:
> 
> Dr. Mangalam;
> Thanks for your help!  I changed my files and they are now working.  A few
> more questions - my experiments are Cy3/Cy5 ratios (so control and
> experimental are from the same chip for each experiment) and I have
> duplicates of each spot per experiment.  

You may already be aware of this, but this is what we call a Paired expt, not Control & Experimental (I know, it's somewhat arbitrary and if you have a better suggestion, we'd be happy to take it).
This is b/c the data are paired and thus have an inherent control for spotting and other errors.


>                                        I have grouped them all together
> under one name as you suggested, so in total I will now have between 6-8
> replicates per file.  I have been using a confidence value for the
> Bayesian prior of 10, does this make sense for the # of replicates?  

Since you're using Paired data, you're not using straight (or transformed) expression values as the input, but a ratio, which cancels the actual expression information and that's what the baysian
prior uses, so you have to calculate a special value in order to make use of the Bayesian estimate.  The rationale for this and the formula to do this is described at:

http://genomics.biochem.uci.edu/genex/cybert/CTHelp.html#PairedExprCol


> I'm
> also wondering if there is some way to distinguish between replicates
> between experiments (ie, points from different chips) and duplicates
> within one experiment (duplicate spots on the same chip), as there is much
> less variance between duplicate spots than between repeated experiments.

Good point.

You're right when you say that there is a difference between on-slide replicates and across-slide repeated measures.  

The statistical purist (or fascist) would say the the on-slide replicates (repeated sampling of 2 or more supposedly identical environments) are meant to be averaged and used as a single number when
comparing across repeated experiments.  The experimenter who paid for the experiment wants brownie points for having done the extra work and isn't willing to give up the extra degrees of freedom.  I
tend to agree with the fascist.  Who wants to carry around extra values if they don't mean anything?

There probably are methods fro doing exactly this, but I don;t know offhand what they are.  It might be interesting to see what teh relative distributions are tho.  Actually, once you are finished
with that data (or if you want to anonymize it - I'm only interestedin the different distributions), I'd be interested in looking at them.

Tony, correct me if this is wrong-headed.



> Thanks,
> Rachel
> 
> ___
> 
> Rachel M. Tompa
> Fred Hutchinson Cancer Research Center
> 1100 Fairview Ave. N.
> A1-111
> Seattle, Washington 98109-1024
> 
> 206-667-4509
> 
> On Thu, 5 Jul 2001, Harry Mangalam wrote:
> 
> > Hi Rachel,
> >
> > From the file you sent, you didn't replace ALL the spaces with '_'.  When you do, it works.  However, if you want xgobi output back, you'll have to rename your duplicate names:
> >
> > ZFP1_promoter -0.032  0.238   0.159   0.368
> > ZFP1_promoter -0.024  0.059   0.171   0.195
> > ^^^^^^^^^^^^^
> >
> > these can't be the same identifier - you're using this approach because you have replicates, no?
> >
> > Then you should separate them out and use them as more replicates, not as different probes with the same name.
> >
> > ie
> > ZFP1_promoter -0.032  0.238   0.159   0.368   -0.024  0.059   0.171   0.195
> >
> > hjm
> >
> >
> >
> > Rachel Tompa wrote:
> > >
> > > I have attached one of my files to this email
> > > Rachel
> > >
> > > ___
> > >
> > > Rachel M. Tompa
> > > Fred Hutchinson Cancer Research Center
> > > 1100 Fairview Ave. N.
> > > A1-111
> > > Seattle, Washington 98109-1024
> > >
> > > 206-667-4509
> > >
> > > On Thu, 5 Jul 2001, Harry Mangalam wrote:
> > >
> > > > Hi Rachel,
> > > >
> > > >
> > > > >From the FAQ:
> > > > (http://genomics.biochem.uci.edu/genex/cybert/CTHelp.html#FAQ55)
> > > >
> > > > 6.What does this error mean?:
> > > >
> > > >     Error in sum(..., na.rm = na.rm) : invalid "mode" of argument
> > > >     Execution halted
> > > >
> > > > It means that a variable was not what R thought it should be. Often
> > > > it is because a label variable had spaces in the name and you
> > > > indicated that the separator should also be whitespace (meaning that
> > > > the numbering of columns gets messed up. First, none of your labels
> > > > should have spaces (or, God forbid, TABS) in them. If you need to use
> > > > a separator, use the '_' character. Second, if this can't be helped,
> > > > try changing the Delimiter from Whitespace (tabs and spaces, the
> > > > default) to TABS only.
> > > >
> > > >
> > > > If this doesn't make sense or it's not applicable, please let me know and I'll try
> > > > to figure it out for you.  If you can send the data file that gives you this error,
> > > > that would help things move
> > > > along.
> > > >
> > > > hjm
> > > >
> > > >
> > > >
> > > >
> > > > Rachel Tompa wrote:
> > > > >
> > > > > Dear Drs. Long and Mangalam;
> > > > > I am trying to run my microarray data on the CyberT website, Paired Data
> > > > > section.  I have files (TAB delimited text files) of 3-4 replicates of
> > > > > experiments, I have already log-transformed the data myself, I have been
> > > > > using the default settings on the website interface, and every time I try
> > > > > to run my data through CyberT I get back this error message:
> > > > >
> > > > > Error in sum(..., na.rm = na.rm) : invalid "mode" of argument
> > > > > Execution halted
> > > > >
> > > > > I have no idea what this message means, can you tell what is wrong with my
> > > > > data format from this?
> > > > > Thank you,
> > > > > Rachel Tompa
> > > > >
> > > > > ___
> > > > >
> > > > > Rachel M. Tompa
> > > > > Fred Hutchinson Cancer Research Center
> > > > > 1100 Fairview Ave. N.
> > > > > A1-111
> > > > > Seattle, Washington 98109-1024
> > > > >
> > > > > 206-667-4509
> > > >
> > > > --
> > > > Cheers,
> > > > Harry
> > > >
> > > > Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...
> > > >
> > >
> > >   --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> > >                       Name: Met1:20all.2
> > >    Met1:20all.2       Type: Plain Text (TEXT/PLAIN)
> > >                   Encoding: BASE64
> > >                Description: CyberT file
> >
> > --
> > Cheers,
> > Harry
> >
> > Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...
> >

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...

[GeneX-dev] Re: CyberT

[Harry M. <man...@ho...>]

Be glad to but no file was attached.  I hate it when I do that... ;)  Friday...

try again.

from the FAQ:
http://genomics.biochem.uci.edu/genex/cybert/CTHelp.html#FAQ55


6.What does this error mean?: 

  Error in sum(..., na.rm = na.rm) : invalid "mode" of argument
  Execution halted
         
It means that a variable was not what R thought it should be. Often it is
because a label variable had spaces in the name and you indicated that the
separator should also be whitespace (meaning that the numbering of columns gets
messed up. First, none of your labels should have spaces (or, God forbid, TABS)
in them. If you need to use a separator, use the '_' character. Second, if this
can't be helped, try changing the Delimiter from Whitespace (tabs and spaces,
the default) to TABS only. 

But it sounds like you have looked at your file pretty carefully.  Nevertheless, of the 50-odd files people have sent to me for debugging, it's always been the case that a space has snuck in where it
wasn't expected or row IDs were missing or were the same (actually the last wouldn;t stop an analysis, but would stop an xgobi visualization..

hjm


Liz Worthey wrote:
> 
> Hi,
>     I just thought that I would drop you a line to see if you had any advice
> about what I could possibly be doing wrong in runing Cyber T. I'm trying to
> use it on a test set of data to examine the levels of variance between a
> number of replicated experiments using a variety of different packages and
> statistical methods. The error message that I get is:
> 
>                 Error in sum(..., na.rm = na.rm) : invalid "mode" of
> argument Execution halted
> 
>     The data is a TAB delimited text file derived from Excel. The first
> column "column 0" is the clone ID and the remaining columns (1-11) are the
> log ratios for the two fluors. Blank instancies where data is lacking for
> some replicates are filled with "NA". Those below background are labelled
> with "0". There are 11 replicates of the same data. The text file looks fine
> to me. I can't detect any TAB or space characters where there shouldn't be
> any.
> 
>     I can't really work out why CyberT isn't working, I have used it before
> successfully. It may be that I am doing something stupid (this has been
> known to occur on a Friday), so I wonder if you might be of some assistance.
> 
>    On the off chance that you might be able to spot something wrong in my
> data. I am attaching the tab delimited file.
> 
>    Thanks in advance for your assistance,
>                                                                      Liz
> 
> Liz Worthey
> Postdoctoral Scientist
> Seattle Biomedical Research Institute
> 206.284.8846. x461
> lwo...@sb...
> 
> 

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...

[GeneX-dev] Re: CyberT Registration and asking a favor

[Harry M. <man...@ho...>]

There is nothing wrong with your file pe se nor with CyberT as far as I can tell.  There is probably  something wrong with your data if these numbers are log-transformed tho (and even if not).  The
values are enormously high and therefore just by looking at them, they will all be significantly different from 'no change'.  I did not even try the Bayesian correction as the data has to be corrected
before you will get anything biologically relevant out of it.

ie
AA211157	7.8472207	8.0578896	8.5579507	7.9703335	8.0476804	8.2401379	-6.680933471

AA563693	7.884906	8.092068	8.1707944	8.0625762	8.0231618	8.074984	-6.749720538

AB002097	7.7775683	8.2664347	8.2236456	7.8408762	7.9888517	8.0304079	-6.779838204


how can your ratios be so high (and tightly grouped)?  Are you sure you didn't make a mistake in transforming your raw data?  If you plot Mean vs the std dev you get a weird distribution (see attached
bad.data.gif (your data) from xgobi).  In a normal distribution (properly tranformed) you should get the familiar '<' shape (see attached good.data.gif)

CyberT reads in and returns correct analysis of these values based on what goes in, but it's "confusion in, confusion out".

harry


"Motoshi Kinouchi, M.D." wrote:
> 
> Hi Harry
> 
> Thank you for your help.
> Maybe I am dense, but I can not run CyberT though following your advice.
> Please help me with running it(Paired data) as I am sending the data file(
> tab-delimited file in Excel) as an attachment.
> Column1-3 are controls, and column7 is Paired Expression Value.
> Numbers of the data are log-transformed ones.
> 
> Thank you again in advance for your help.
> 
> Smile & Mercy,
> MOTOSHI
> --                                  Motoshi Kinouchi, M.D
>              Department of Dermatology, Boston University School of Medicine
>                                 609 Albany st. Boston, MA 02118
>                                      Tel: 629-638-5559
>                                      Fax: 617-414-1359
> 
> > From: Harry Mangalam <man...@ho...>
> > Date: Tue, 03 Jul 2001 06:46:21 -0700
> > To: "Motoshi Kinouchi, M.D." <kin...@bu...>, genexdev at SF
> > <gen...@li...>
> > Subject: Re: CyberT Registration and asking a favor
> >
> > Hi Motoshi,  Sorry for the dealy, I was out of town last week.
> >
> >
> >> From the FAQ:
> > (http://genomics.biochem.uci.edu/genex/cybert/CTHelp.html#FAQ55)
> >
> > 6.What does this error mean?:
> >
> > Error in sum(..., na.rm = na.rm) : invalid "mode" of argument
> > Execution halted
> >
> > It means that a variable was not what R thought it should be. Often
> > it is because a label variable had spaces in the name and you
> > indicated that the separator should also be whitespace (meaning that
> > the numbering of columns gets messed up. First, none of your labels
> > should have spaces (or, God forbid, TABS) in them. If you need to use
> > a separator, use the '_' character. Second, if this can't be helped,
> > try changing the Delimiter from Whitespace (tabs and spaces, the
> > default) to TABS only.
> >
> >
> > If this doesn't make sense or it's not applicable, please let me know and I'll
> > try to figure it out for you.  If you can send the data fiel that gives you
> > this error, that would help things move
> > along.
> >
> > hjm
> >
> >
> >
> > "Motoshi Kinouchi, M.D." wrote:
> >>
> >> Hi
> >>
> >> CyberT looks great. However, I can not run it well.
> >> When I submit data, I get the following message.
> >>
> >> Error in 1:row.lens[1] : NA/NaN argument
> >> Execution halted
> >>
> >> As I am not familiar with R, I do not understand what it means.
> >> Please tell me the solution easily.
> >>
> >> Thank you for your time in this matter.
> >>
> >> Motoshi
> >> --                                  Motoshi Kinouchi, M.D
> >> Department of Dermatology, Boston University School of Medicine
> >> 609 Albany st. Boston, MA 02118
> >> Tel: 629-638-5559
> >> Fax: 617-414-1359
> >
> > --
> > Cheers,
> > Harry
> >
> > Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...
> >
> >
> >
> 
>   --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>                                  Name: st.age.uvb for cyter-t.tab
>    st.age.uvb for cyter-t.tab    Type: application/x-msexcel
>                              Encoding: base64

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...

Re: [GeneX-dev] GENE SEQUENCING?

[Harry M. <man...@ho...>]

Nope, you want Phred/Phrap if you can work on Unix/Linux or a commercial product like Sequencher from GeneCodes.  

GeneX is a gene expression database system.

> "Claris Castillo T." wrote:
> 
> I am interesting in doing  some "gene sequencing". Does your software that?
> I am working in  a project for the Smithsonian Tropical Research Institue, at Panama City, Panama.
> Thanks,
> Claris

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...

[GeneX-dev] GENE SEQUENCING?

[Claris C. T. <cla...@cw...>]

I am interesting in doing  some "gene sequencing". Does your software =
that?
I am working in  a project for the Smithsonian Tropical Research =
Institue, at Panama City, Panama.
Thanks,
Claris

[GeneX-dev] Re: CyberT Registration and asking a favor

[Harry M. <man...@ho...>]

Hi Motoshi,  Sorry for the dealy, I was out of town last week.


>From the FAQ:
(http://genomics.biochem.uci.edu/genex/cybert/CTHelp.html#FAQ55)

6.What does this error mean?: 

    Error in sum(..., na.rm = na.rm) : invalid "mode" of argument
    Execution halted
         
It means that a variable was not what R thought it should be. Often
it is because a label variable had spaces in the name and you
indicated that the separator should also be whitespace (meaning that
the numbering of columns gets messed up. First, none of your labels
should have spaces (or, God forbid, TABS) in them. If you need to use
a separator, use the '_' character. Second, if this can't be helped,
try changing the Delimiter from Whitespace (tabs and spaces, the
default) to TABS only. 


If this doesn't make sense or it's not applicable, please let me know and I'll try to figure it out for you.  If you can send the data fiel that gives you this error, that would help things move
along.

hjm



"Motoshi Kinouchi, M.D." wrote:
> 
> Hi
> 
> CyberT looks great. However, I can not run it well.
> When I submit data, I get the following message.
> 
> Error in 1:row.lens[1] : NA/NaN argument
> Execution halted
> 
> As I am not familiar with R, I do not understand what it means.
> Please tell me the solution easily.
> 
> Thank you for your time in this matter.
> 
> Motoshi
> --                                  Motoshi Kinouchi, M.D
>              Department of Dermatology, Boston University School of Medicine
>                                 609 Albany st. Boston, MA 02118
>                                      Tel: 629-638-5559
>                                      Fax: 617-414-1359

-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...