Re: Segmentation Fault - fixed

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: Segmentation Fault - fixed

Benjamin Hitz

I just wanted to report that this patch did indeed fix my augustus on mac os X.   It actually finished weeks ago, but I haven't had time to look at the results until this week.

Thanks,
Ben
On Feb 4, 2011, at 1:48 AM, Mario Stanke wrote:

Yes, Rodolfo, just replace and recompile. I will also make a new augustus release with the bugfix and other updates in the next few days.
Thanks for your offer, I may get back to you on that.

Peace, Mario

2011/2/4 Rodolfo Aramayo <[hidden email]>
Thanks Mario and Carson and now I suppose the only thing to do is to replace that file and recompile??

And in the future Mario, if you need access to a Mac machine please contact me and I will gladly give you access to one of my machines

Viva Augustus!

Cheers,

--Rodolfo





On Thu, Feb 3, 2011 at 14:56, Mario Stanke <[hidden email]> wrote:
Thanks. This was an bug I must have introduced not much more than a year ago. It may just not show up on some systems, but I am glad you reproduced it and we found it.

Thanks, Mario


2011/2/3 Carson Holt <[hidden email]>
It worked all the way through on the test sequence.

Thanks,
Carson



On 2/3/11 11:41 AM, "Mario Stanke" <mario.stanke@...> wrote:

Thanks.
Please try whether it works with the attached patched extrinsicinfo.cc.
Mario


2011/2/3 Carson Holt <carson.holt@...>
Here is what I get.

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x00000001020a1fff
SequenceFeatureCollection::prepareLocalMalus (this=0x7fff5fbfd590, dna=<value temporarily unavailable, due to optimizations>) at geneticcode.hh:147
147            erg |= base2int(wcComplement(s[i])) << (2*i);
(gdb) bt
#0  SequenceFeatureCollection::prepareLocalMalus (this=0x7fff5fbfd590, dna=<value temporarily unavailable, due to optimizations>) at geneticcode.hh:147
#1  0x0000000100046dbf in StateModel::setSFC () at /usr/local/maker/exe/augustus/include/statemodel.hh:679
#2  0x0000000100046dbf in NAMGene::getStepGenes (this=0x7fff5fbfda20, annoseq=<value temporarily unavailable, due to optimizations>, sfc=@0x7fff5fbfd590, strand=bothstrands, onlyViterbi=true) at namgene.cc:681
#3  0x0000000100049503 in NAMGene::doViterbiPiecewise (this=0x7fff5fbfda20, sfc=@0x10020c760, annoseq=0x100208870, strand=bothstrands) at namgene.cc:585
#4  0x0000000100006bf3 in predictOnInputSequences () at iostream:77
#5  0x0000000100007660 in main () at iostream:77
Current language:  auto; currently c++


Thanks,
Carson


On 2/3/11 7:08 AM, "Mario Stanke" <mario.stanke@... <http://mario.stanke@...> > wrote:

Hi Rodolfo, Carson, Benjamin,

unfortunately, I don't have a Mac at hand to reproduce the error. On my machine, 64 bit Ubuntu, the
command
augustus --species=neurospora scontig1011.07_wt.fasta
runs through without error.

As I can't find the error without reproducing it, could you do me a favor and do the following on a machine where it fails:
Edit the src/Makefile so that it says

CFLAGS := -Wall -Wno-sign-compare -ansi -pedantic -O2 ${CFLAGS} -g -ggdb

Then do a
make clean all
in the src directory and
run the command that causes the segmentation fault with the debugger gdb (or another debugger):

gdb augustus
set args --species=neurospora scontig1011.07_wt.fasta
r

When it terminates (hopefully with the same segmentation fault), type
bt
This returns the list of nested function calls. Then at least we could have a clue where in the code a problem may be.

Best, Mario



2011/1/31 Rodolfo Aramayo <raramayo@... <http://raramayo@...> >
If this is correct then we should be able to find a fungal sequence that does not induce the error, right?

Similarly, we should be able to run a, say Drosophila sequence, using the fungal HMM  without a glitch, right?

I will test both these two assertions and get back to you

If you have a test seq, please send it in my direction.


--Rodolfo





On Mon, Jan 31, 2011 at 12:12, Carson Holt <carson.holt@... <http://carson.holt@...> > wrote:
I still get the failure using human HMM on the fungal sequence.   Sometimes specific sequences can bring out errors because the patterns within causes the exact set of sequential events needed to trigger a bug.  For example I get a RepeatMasker failure on a specific contig in lamprey that I can’t duplicate with any other species or sequence.  It ended up being a perfect pattern to cause certain events to occur improperly.

--Carson


On 1/31/11 11:04 AM, "Rodolfo Aramayo" <raramayo@... <http://raramayo@...>  <http://raramayo@...> > wrote:

I did not use Apple's compilers

I used gnuGCC

The problem is real in both x86_64 and i386 Mac machines

I find it hard to believe that a fasta file can be the culprit

I think the problem is with the fungal-associated HMMs

Look at them, they are different from the human ones

I think there lies the problem...


--Rodolfo




On Mon, Jan 31, 2011 at 11:53, Carson Holt <carson.holt@... <http://carson.holt@...>  <http://carson.holt@...> > wrote:
I ran the same sequence Rodolfo attached in his last e-mail (augustus.zip).  I ran it with multiple species (human, elegan, drosophila, and fungi species).  It seems to be something in the sequence causing the error.

One of the difficulties with the latest OSX is that it is a hybrid OS, some parts are compiled 32 bit while others are 64 bit.  This affects libraries and causes weird behavior on many linux based programs on new Macs.  Perhaps a different compiler than xcode g++ that comes with Mac could help, but I have no good suggestions.

Thanks,
Carson


On 1/31/11 10:31 AM, "Benjamin Hitz" <hitz@... <http://hitz@...>  <http://hitz@...>  <http://hitz@...> > wrote:


Do you guys have a specific use-case where it fails (given inputs, etc.)?  Perhaps in combination with my cases we can figure it out ourselves.  Were you running it against cerevisiae?  Perhaps that particular HMM is "bad".
Do you remember how you compiled augustus?   I guess the instructions are just to "make", but perhaps there are some compile flags we could set, run it through gdb or something.

I'll report it though,
Ben


On Jan 31, 2011, at 8:27 AM, Carson Holt wrote:

Rodolfo is correct.  There seems to be a bug in augustus.  It fails only on Mac OSX on some sequences.  We submited a bug report to them several days ago, but It might be a good idea fork you to submit a bug report to them as well.

Thanks,
Carson


On 1/29/11 5:45 PM, "Rodolfo Aramayo" <raramayo@... <http://raramayo@...>  <http://raramayo@...>  <http://raramayo@...>  <x-msg:<a href="x-msg://13/raramayo@tamu.edu" target="_blank">//13/raramayo@...> > wrote:


--Rodolfo Aramayo








---------- Forwarded message ----------
From: Rodolfo Aramayo <raramayo@... <http://raramayo@...>  <http://raramayo@...>  <http://raramayo@...>  <x-msg:<a href="x-msg://13/raramayo@tamu.edu" target="_blank">//13/raramayo@...> >
Date: Mon, Jan 3, 2011 at 17:29
Subject: Re: Segmentation Fault
To: "mstanke@... <http://mstanke@...>  <http://mstanke@...>  <http://mstanke@...>  <x-msg:<a href="x-msg://13/mstanke@gwdg.de" target="_blank">//13/mstanke@...> " <mstanke@... <http://mstanke@...>  <http://mstanke@...>  <http://mstanke@...>  <x-msg:<a href="x-msg://13/mstanke@gwdg.de" target="_blank">//13/mstanke@...> >
Cc: Carson Holt <carson.holt@... <http://carson.holt@...>  <http://carson.holt@...>  <http://carson.holt@...>  <x-msg:<a href="x-msg://13/carson.holt@genetics.utah.edu" target="_blank">//13/carson.holt@...> >


Hi,

Thanks Carson.

I should also add that I have tried both version 2.4 and 2.5 in two separate computers both running MacOS:

ProductName:    Mac OS X Server
ProductVersion:    10.6.5
BuildVersion:    10H575

One is a "i386" and the other a "x86_64" machine

Both Augustus2.4 and 2.5 were compiled and tested on those machines according to the protocol described below

In all cases Augustus failed

Thanks

--Rodolfo Aramayo


==
INSTALLING AUGUSTUS
http://augustus.gobics.de/

>sudo su
>wget http://augustus.gobics.de/binaries/augustus.2.5.tar.gz
>gnutar -xzvf augustus.2.5.tar.gz
>cd augustus.2.5
>find . -type d -exec chown -v root:wheel {} \;
>find . -type f -exec chown -v root:wheel {} \;
>cd src/
>make
>cd ../../
>ln -s augustus.2.5/ augustus

==
=>add to /etc/profile
#==========
#TO_AUGUSTUS (/usr/local/bioinfosoft/augustus):SERVER_AND_CLIENT
export PATH=/usr/local/bioinfosoft/augustus/bin:$PATH
export AUGUSTUS_CONFIG_PATH=/usr/local/bioinfosoft/augustus/config/
export PATH=/usr/local/bioinfosoft/augustus/scripts:$PATH
export PATH=/usr/local/bioinfosoft/augustus/config/species:$PATH
#==========
==





On Mon, Jan 3, 2011 at 17:16, Carson Holt <carson.holt@... <http://carson.holt@...>  <http://carson.holt@...>  <http://carson.holt@...>  <x-msg:<a href="x-msg://13/carson.holt@genetics.utah.edu" target="_blank">//13/carson.holt@...> > wrote:
Hello both myself and a user I am helping have encountered a bug in Augustus on separate Mac OSX machines.  We get a segmentation fault when predicting on the attached contig regardless of what species is selected.  The correct species is neurospora.

On Linux, Augustus performs just fine on the contig and completes all predictions, so this is Mac OSX specific issue.  The failure occurs for me on Augustus version 2.3 as well as the more up to date 2.5.

Thank you,
Carson Holt




--
Ben Hitz
Senior Scientific Programmer ** Saccharomyces Genome Database ** GO Consortium
Stanford University ** hitz@... <http://hitz@...>  <http://hitz@...>  <http://hitz@...>














--
----------------------------------------------------
Prof. Dr. Mario Stanke
Universität Greifswald
Institut für Mathematik und Informatik
Walther-Rathenau-Straße 47
17487 Greifswald
Germany
+49 3834 864642




--
----------------------------------------------------
Prof. Dr. Mario Stanke
Universität Greifswald
Institut für Mathematik und Informatik
Walther-Rathenau-Straße 47
17487 Greifswald
Germany
+49 3834 864642

--
Ben Hitz 
Senior Scientific Programmer ** Saccharomyces Genome Database ** GO Consortium
Stanford University ** [hidden email]




_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Reply | Threaded
Open this post in threaded view
|

Re: Segmentation Fault - fixed

Rodolfo Aramayo
Yes, me too. I was able to re-annotate the Neurospora genome
generating Augustus predictions. Pretty cool

--Rodolfo





On Tue, Mar 1, 2011 at 12:09, Benjamin Hitz <[hidden email]> wrote:

>
> I just wanted to report that this patch did indeed fix my augustus on mac os
> X.   It actually finished weeks ago, but I haven't had time to look at the
> results until this week.
> Thanks,
> Ben
> On Feb 4, 2011, at 1:48 AM, Mario Stanke wrote:
>
> Yes, Rodolfo, just replace and recompile. I will also make a new augustus
> release with the bugfix and other updates in the next few days.
> Thanks for your offer, I may get back to you on that.
>
> Peace, Mario
>
> 2011/2/4 Rodolfo Aramayo <[hidden email]>
>>
>> Thanks Mario and Carson and now I suppose the only thing to do is to
>> replace that file and recompile??
>>
>> And in the future Mario, if you need access to a Mac machine please
>> contact me and I will gladly give you access to one of my machines
>>
>> Viva Augustus!
>>
>> Cheers,
>>
>> --Rodolfo
>>
>>
>>
>>
>> On Thu, Feb 3, 2011 at 14:56, Mario Stanke <[hidden email]> wrote:
>>>
>>> Thanks. This was an bug I must have introduced not much more than a year
>>> ago. It may just not show up on some systems, but I am glad you reproduced
>>> it and we found it.
>>>
>>> Thanks, Mario
>>>
>>> 2011/2/3 Carson Holt <[hidden email]>
>>>>
>>>> It worked all the way through on the test sequence.
>>>>
>>>> Thanks,
>>>> Carson
>>>>
>>>>
>>>> On 2/3/11 11:41 AM, "Mario Stanke" <[hidden email]> wrote:
>>>>
>>>> Thanks.
>>>> Please try whether it works with the attached patched extrinsicinfo.cc.
>>>> Mario
>>>>
>>>>
>>>> 2011/2/3 Carson Holt <[hidden email]>
>>>>
>>>> Here is what I get.
>>>>
>>>> Program received signal EXC_BAD_ACCESS, Could not access memory.
>>>> Reason: KERN_INVALID_ADDRESS at address: 0x00000001020a1fff
>>>> SequenceFeatureCollection::prepareLocalMalus (this=0x7fff5fbfd590,
>>>> dna=<value temporarily unavailable, due to optimizations>) at
>>>> geneticcode.hh:147
>>>> 147            erg |= base2int(wcComplement(s[i])) << (2*i);
>>>> (gdb) bt
>>>> #0  SequenceFeatureCollection::prepareLocalMalus (this=0x7fff5fbfd590,
>>>> dna=<value temporarily unavailable, due to optimizations>) at
>>>> geneticcode.hh:147
>>>> #1  0x0000000100046dbf in StateModel::setSFC () at
>>>> /usr/local/maker/exe/augustus/include/statemodel.hh:679
>>>> #2  0x0000000100046dbf in NAMGene::getStepGenes (this=0x7fff5fbfda20,
>>>> annoseq=<value temporarily unavailable, due to optimizations>,
>>>> sfc=@0x7fff5fbfd590, strand=bothstrands, onlyViterbi=true) at namgene.cc:681
>>>> #3  0x0000000100049503 in NAMGene::doViterbiPiecewise
>>>> (this=0x7fff5fbfda20, sfc=@0x10020c760, annoseq=0x100208870,
>>>> strand=bothstrands) at namgene.cc:585
>>>> #4  0x0000000100006bf3 in predictOnInputSequences () at iostream:77
>>>> #5  0x0000000100007660 in main () at iostream:77
>>>> Current language:  auto; currently c++
>>>>
>>>>
>>>> Thanks,
>>>> Carson
>>>>
>>>>
>>>> On 2/3/11 7:08 AM, "Mario Stanke" <[hidden email]
>>>> <http://mario.stanke@...> > wrote:
>>>>
>>>> Hi Rodolfo, Carson, Benjamin,
>>>>
>>>> unfortunately, I don't have a Mac at hand to reproduce the error. On my
>>>> machine, 64 bit Ubuntu, the
>>>> command
>>>> augustus --species=neurospora scontig1011.07_wt.fasta
>>>> runs through without error.
>>>>
>>>> As I can't find the error without reproducing it, could you do me a
>>>> favor and do the following on a machine where it fails:
>>>> Edit the src/Makefile so that it says
>>>>
>>>> CFLAGS := -Wall -Wno-sign-compare -ansi -pedantic -O2 ${CFLAGS} -g -ggdb
>>>>
>>>> Then do a
>>>> make clean all
>>>> in the src directory and
>>>> run the command that causes the segmentation fault with the debugger gdb
>>>> (or another debugger):
>>>>
>>>> gdb augustus
>>>> set args --species=neurospora scontig1011.07_wt.fasta
>>>> r
>>>>
>>>> When it terminates (hopefully with the same segmentation fault), type
>>>> bt
>>>> This returns the list of nested function calls. Then at least we could
>>>> have a clue where in the code a problem may be.
>>>>
>>>> Best, Mario
>>>>
>>>>
>>>>
>>>> 2011/1/31 Rodolfo Aramayo <[hidden email] <http://raramayo@...>
>>>> >
>>>>
>>>> If this is correct then we should be able to find a fungal sequence that
>>>> does not induce the error, right?
>>>>
>>>> Similarly, we should be able to run a, say Drosophila sequence, using
>>>> the fungal HMM  without a glitch, right?
>>>>
>>>> I will test both these two assertions and get back to you
>>>>
>>>> If you have a test seq, please send it in my direction.
>>>>
>>>>
>>>> --Rodolfo
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Mon, Jan 31, 2011 at 12:12, Carson Holt
>>>> <[hidden email] <http://carson.holt@...> >
>>>> wrote:
>>>>
>>>> I still get the failure using human HMM on the fungal sequence.
>>>>   Sometimes specific sequences can bring out errors because the patterns
>>>> within causes the exact set of sequential events needed to trigger a bug.
>>>>  For example I get a RepeatMasker failure on a specific contig in lamprey
>>>> that I can’t duplicate with any other species or sequence.  It ended up
>>>> being a perfect pattern to cause certain events to occur improperly.
>>>>
>>>> --Carson
>>>>
>>>>
>>>> On 1/31/11 11:04 AM, "Rodolfo Aramayo" <[hidden email]
>>>> <http://raramayo@...>  <http://raramayo@...> > wrote:
>>>>
>>>> I did not use Apple's compilers
>>>>
>>>> I used gnuGCC
>>>>
>>>> The problem is real in both x86_64 and i386 Mac machines
>>>>
>>>> I find it hard to believe that a fasta file can be the culprit
>>>>
>>>> I think the problem is with the fungal-associated HMMs
>>>>
>>>> Look at them, they are different from the human ones
>>>>
>>>> I think there lies the problem...
>>>>
>>>>
>>>> --Rodolfo
>>>>
>>>>
>>>>
>>>>
>>>> On Mon, Jan 31, 2011 at 11:53, Carson Holt
>>>> <[hidden email] <http://carson.holt@...>
>>>>  <http://carson.holt@...> > wrote:
>>>>
>>>> I ran the same sequence Rodolfo attached in his last e-mail
>>>> (augustus.zip).  I ran it with multiple species (human, elegan, drosophila,
>>>> and fungi species).  It seems to be something in the sequence causing the
>>>> error.
>>>>
>>>> One of the difficulties with the latest OSX is that it is a hybrid OS,
>>>> some parts are compiled 32 bit while others are 64 bit.  This affects
>>>> libraries and causes weird behavior on many linux based programs on new
>>>> Macs.  Perhaps a different compiler than xcode g++ that comes with Mac could
>>>> help, but I have no good suggestions.
>>>>
>>>> Thanks,
>>>> Carson
>>>>
>>>>
>>>> On 1/31/11 10:31 AM, "Benjamin Hitz" <[hidden email]
>>>> <http://hitz@...>  <http://hitz@...>
>>>>  <http://hitz@...> > wrote:
>>>>
>>>>
>>>> Do you guys have a specific use-case where it fails (given inputs,
>>>> etc.)?  Perhaps in combination with my cases we can figure it out ourselves.
>>>>  Were you running it against cerevisiae?  Perhaps that particular HMM is
>>>> "bad".
>>>> Do you remember how you compiled augustus?   I guess the instructions
>>>> are just to "make", but perhaps there are some compile flags we could set,
>>>> run it through gdb or something.
>>>>
>>>> I'll report it though,
>>>> Ben
>>>>
>>>>
>>>> On Jan 31, 2011, at 8:27 AM, Carson Holt wrote:
>>>>
>>>> Rodolfo is correct.  There seems to be a bug in augustus.  It fails only
>>>> on Mac OSX on some sequences.  We submited a bug report to them several days
>>>> ago, but It might be a good idea fork you to submit a bug report to them as
>>>> well.
>>>>
>>>> Thanks,
>>>> Carson
>>>>
>>>>
>>>> On 1/29/11 5:45 PM, "Rodolfo Aramayo" <[hidden email]
>>>> <http://raramayo@...>  <http://raramayo@...>
>>>>  <http://raramayo@...>  <x-msg://13/[hidden email]> > wrote:
>>>>
>>>>
>>>> --Rodolfo Aramayo
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> ---------- Forwarded message ----------
>>>> From: Rodolfo Aramayo <[hidden email] <http://raramayo@...>
>>>>  <http://raramayo@...>  <http://raramayo@...>
>>>>  <x-msg://13/[hidden email]> >
>>>> Date: Mon, Jan 3, 2011 at 17:29
>>>> Subject: Re: Segmentation Fault
>>>> To: "[hidden email] <http://mstanke@...>  <http://mstanke@...>
>>>>  <http://mstanke@...>  <x-msg://13/[hidden email]> " <[hidden email]
>>>> <http://mstanke@...>  <http://mstanke@...>  <http://mstanke@...>
>>>>  <x-msg://13/[hidden email]> >
>>>> Cc: Carson Holt <[hidden email]
>>>> <http://carson.holt@...>
>>>>  <http://carson.holt@...>
>>>>  <http://carson.holt@...>
>>>>  <x-msg://13/[hidden email]> >
>>>>
>>>>
>>>> Hi,
>>>>
>>>> Thanks Carson.
>>>>
>>>> I should also add that I have tried both version 2.4 and 2.5 in two
>>>> separate computers both running MacOS:
>>>>
>>>> ProductName:    Mac OS X Server
>>>> ProductVersion:    10.6.5
>>>> BuildVersion:    10H575
>>>>
>>>> One is a "i386" and the other a "x86_64" machine
>>>>
>>>> Both Augustus2.4 and 2.5 were compiled and tested on those machines
>>>> according to the protocol described below
>>>>
>>>> In all cases Augustus failed
>>>>
>>>> Thanks
>>>>
>>>> --Rodolfo Aramayo
>>>>
>>>>
>>>> ==
>>>> INSTALLING AUGUSTUS
>>>> http://augustus.gobics.de/
>>>>
>>>> >sudo su
>>>> >wget http://augustus.gobics.de/binaries/augustus.2.5.tar.gz
>>>> >gnutar -xzvf augustus.2.5.tar.gz
>>>> >cd augustus.2.5
>>>> >find . -type d -exec chown -v root:wheel {} \;
>>>> >find . -type f -exec chown -v root:wheel {} \;
>>>> >cd src/
>>>> >make
>>>> >cd ../../
>>>> >ln -s augustus.2.5/ augustus
>>>>
>>>> ==
>>>> =>add to /etc/profile
>>>> #==========
>>>> #TO_AUGUSTUS (/usr/local/bioinfosoft/augustus):SERVER_AND_CLIENT
>>>> export PATH=/usr/local/bioinfosoft/augustus/bin:$PATH
>>>> export AUGUSTUS_CONFIG_PATH=/usr/local/bioinfosoft/augustus/config/
>>>> export PATH=/usr/local/bioinfosoft/augustus/scripts:$PATH
>>>> export PATH=/usr/local/bioinfosoft/augustus/config/species:$PATH
>>>> #==========
>>>> ==
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Mon, Jan 3, 2011 at 17:16, Carson Holt <[hidden email]
>>>> <http://carson.holt@...>
>>>>  <http://carson.holt@...>
>>>>  <http://carson.holt@...>
>>>>  <x-msg://13/[hidden email]> > wrote:
>>>> Hello both myself and a user I am helping have encountered a bug in
>>>> Augustus on separate Mac OSX machines.  We get a segmentation fault when
>>>> predicting on the attached contig regardless of what species is selected.
>>>>  The correct species is neurospora.
>>>>
>>>> On Linux, Augustus performs just fine on the contig and completes all
>>>> predictions, so this is Mac OSX specific issue.  The failure occurs for me
>>>> on Augustus version 2.3 as well as the more up to date 2.5.
>>>>
>>>> Thank you,
>>>> Carson Holt
>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Ben Hitz
>>>> Senior Scientific Programmer ** Saccharomyces Genome Database ** GO
>>>> Consortium
>>>> Stanford University ** [hidden email]
>>>> <http://hitz@...>  <http://hitz@...>
>>>>  <http://hitz@...>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>> --
>>> ----------------------------------------------------
>>> Prof. Dr. Mario Stanke
>>> Universität Greifswald
>>> Institut für Mathematik und Informatik
>>> Walther-Rathenau-Straße 47
>>> 17487 Greifswald
>>> Germany
>>> +49 3834 864642
>>
>
>
>
> --
> ----------------------------------------------------
> Prof. Dr. Mario Stanke
> Universität Greifswald
> Institut für Mathematik und Informatik
> Walther-Rathenau-Straße 47
> 17487 Greifswald
> Germany
> +49 3834 864642
>
> --
> Ben Hitz
> Senior Scientific Programmer ** Saccharomyces Genome Database ** GO
> Consortium
> Stanford University ** [hidden email]
>
>
>

_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org