modelseed / model-seed-core Goto Github PK

I hate having to commit to chris's repository to make changes. I've created a ModelSEED organization.

Chris, you need to move your repo over to the organization and fork a new copy from it. This way you'll have your own copy with your own master back.

OK, I still got the DB error:

Failure running
/Users/seaver/Software/Model-SEED-core/lib/ModelSEED/ModelDriver.pl:
DBI connect('database=ModelDB;host=localhost;port=3306;mysql_socket=/var/lib/mysql/mysql.sock','ModelSEED',...)
failed: Can't connect to local MySQL server through socket '/var/lib/mysql/mysql.sock' (2) at /Users/seaver/Software/Model-SEED-core/lib/PPO/PPOBackend/MySQL.pm
line 82

As it turns out, the code in lib/PPO/PPOBackend/MySQL.pm is genuinely trying to connect through the socket at /var/lib/mysql/mysql.sock whereas on my mac, the MySQL installation runs the socket at /tmp/mysql.sock

so this command:

sudo ln -s /tmp/mysql.sock /var/lib/mysql/mysql.sock

fixed the problem, so that's probably another line needed for the instructions, that the code assumes the socket is placed at /var/lib/mysql, and a soft link may be needed to fix this.

After setting the the parameters in the setting under config.

After running the script I get:

raphyz@laptopraphyz ~/phd/ModelSEED/bin $ ./configureModelSEED.sh 
mkdir /config: Permission denied at ../lib/ModelSEED/ModelSEEDScripts/configureModelSEED.pl line 71

I have also tried it when pointing to the config file (I saw in the perl that you expect parameters although in the install you did not mention them.

Remember that I am not running as a root user and as such I do not have permissions to create folders under /.
Submitted by Raphy Zareki

The first problem with this scirpt is that you need to clone the repository in order to see it, and it clones the repository with it.

The git clone command does not work for me for some reason, and I had to change the git clone command to: git clone https://[email protected]/ModelSEED/Model-SEED-core $INSTALL_DIR;

where user_name was changed to my user name.

I wonder if this will only be my problem.

I get the following error when running: CreateDBScheme.sh

PPO Generator:
XML definition successfully read.
[MySQL.pm 85]Transaction aborted: DBD::mysql::db do failed: Can't create database 'ModelDB'; database exists at /home/raphyz/Desktop/SHARE/phd/ModelSEED/lib/PPO/PPOBackend/MySQL.pm line 78.
Creating database for PPO... failed.
.. with Unable to connect to database backend. at /home/raphyz/Desktop/SHARE/phd/ModelSEED/lib/PPO/PPOGenerator.pm line 717.

It is probably related to the default charset defined on the machine, and the charset created when opening the connection with the database.

Ok, to get the unit tests working again we need something like DevModelDB except local to the core.

I'm thinking we package a small sqlite database ( one or two models, two dummy users, compounds, rxns, media, etc. ) in a build script along with the /data/models for those two models.

TestingHelpers.pm would download these files to /test/ if they weren't already there. On a test startup, these would be copied to tmp directories and used. I think this will be faster (and more portable) than the DevModelDB approach.

So, I've installed the KEGG and MetaCyc databases as models, with their own private databases. However, when I install MetaCyc, I use the KEGG private database thus:

bash ModelDriver.sh importmodel?MetaCyc?83333.1?seaver?/Users/seaver/Documents/Projects/Model_Import_Files/?1?/Users/seaver/Software/Model-SEED-core/models/seaver/KEGGimport.2/0/biochemistry/

This runs and completes. However, an analysis of the output shows that the process for MetaCyc didn't completely use the KEGG database.

1. When KEGG installed, it found new compounds and reactions thus:

New:cpd80000 for C06816
New:rxn80000 for R00001

When MetaCyc installed, it found new compounds and reactions thus:

New:cpd80000 for 28-Homocastasterone-Sulfates
New:rxn80000 for CHONDROITIN-4-SULFOTRANSFERASE-RXN

Given that it's using the KEGG private metabolic database and we're working towards the reconcilation of metabolic models, the first temporary compound and reaction id for the MetaCyc database should be the last one for the KEGG database, plus one.

2. The installation process for MetaCyc did not find any compounds that were in the new KEGG database, and were not already in the initial biochemistry database. In other words, it did not find any new compounds with a temporary id. There's a chance that this is not a bug, but given the size of the two databases, I doubt it.

However, it did find reactions with a temporary id from the new KEGG database. ie:

Found:rxn80428 for RXN-12014

So the code below does not work in windows for a variety of reasons. I am proposing we include ModelDB.db in the test data directory instead of the sql file. Would this cause problems? I also suggest adding the test directory to the standard flatfile install so it's immediately in place for tests to be performed on a fresh installation. We'll need some way of indicating when an data update would be run. This is something to think about.

TestingHelper.pm: line 182...
if ( !-d $TestDataDir ) {
# If we don't have the database try to download it
# Fail if we are still unable to get the database.
if ( !-f "$dataDir/TestDB.tgz") {
system("curl $TestDbURL 2> /dev/null > $dataDir/TestDB.tgz");#Note, this does not work in windows...
if (!-f "$dataDir/TestDB.tgz") {
ModelSEED::FIGMODEL::FIGMODELERROR("Unable to copy TestDB.tgz from $TestDbURL");
}
}
mkdir $TestDataDir;
system("tar -xzf $dataDir/TestDB.tgz -C $TestDataDir");#Note, this does not work in windows...
if( !-d "$TestDataDir/data/ModelDB" || !-f "$TestDataDir/data/ModelDB/ModelDB.sqlite") {
ModelSEED::FIGMODEL::FIGMODELERROR("TestDB.tgz does not look like I expected!");
}
system("sqlite3 $TestDataDir/data/ModelDB/ModelDB.db < $TestDataDir/data/ModelDB/ModelDB.sqlite");
}

When I load the ms config file (bin directory, bioenvironment):

[jplfaria@biologin-4 bin]$ ./ms-config load

I run into this error:

g++ -O3 -fPIC -fexceptions -DNDEBUG -DIL_STD -DILOSTRICTPOD -DLINUX -I/vol/model-prod/Software/solvers/cplex/include/ilcplex/ -I/vol/model-prod/Software/solvers/glpk-4.38/include -I../Include -DNOSAFEMEM -DNOBLOCKMEM -o mfatoolkit ../Source/driver.o ../Source/MFAProblem.o ../Source/CPLEXapi.o ../Source/SCIPapi.o ../Source/GLPKapi.o ../Source/LINDOapiEMPTY.o ../Source/SolverInterface.o ../Source/Species.o ../Source/Data.o ../Source/InterfaceFunctions.o ../Source/Identity.o ../Source/Reaction.o ../Source/GlobalFunctions.o ../Source/AtomCPP.o ../Source/UtilityFunctions.o ../Source/AtomType.o ../Source/Gene.o ../Source/GeneInterval.o ../Source/stringDB.o -L/vol/model-prod/Software/solvers/glpk-4.38/src/.libs -lglpk -L/home/chenry/Software/CPLEX_Studio_AcademicResearch122/cplex/lib/x86-64_sles10_4.1/static_pic -lcplex -lm -lpthread -lz -framework CoreFoundation -framework IOKit
g++: CoreFoundation: No such file or directory
g++: IOKit: No such file or directory
make: *** [MFATOOLKIT] Error 1

Those entires in the make file should only appear when cplex is being used on OSX. Otherwise, they should be removed.

José

The perl script is a relative path which means it's called with the cwd, this might not bin.

$ pwd
/Users/devoid/Desktop/core-test
$ ./bin/configureModelSEED.sh
Can't open perl script "../lib/ModelSEED/ModelSEEDScripts/configureModelSEED.pl": No such file or directory

Well I went through the process twice.
Since I am not an "advanced user" I had some troubles that I manage to solve but I am seeing a lot of people having the same problems and not being able to solve them.
List of issues I found:
-SQLite database merge issue (described on the previous issue here on GitHub).
-More Perl Libs missing than the ones on the list:
Config::Tiny
LWP::Simple
DBI
-GLPK, can be difficult to set up, can you distribute together with the ModelSEEDcore?

Ultimately, having an installer distributing the SQLdb, The Perl libs and the open source Solvers like GLPK.

José

It would be nice if the command lines would have a usage message for example:

./importmodel --help

or -?

I had to guess that for importmodel the -overwrite parameter expects a 0|1 and not yes|no or something else
the usage messge should list that

I was wondering whether there's a way to see where the files I load to the database from MSModelFiles go? I wanted to keep track of the model/biomass files I personally have loaded and to delete the files I no longer want there (I've been uploading a lot of random files where I was just trying out stuff).

Found:rxn00806 for LEUTAi
Found:rxn00902 for IPPS
Found:rxn03062 for IPMD
Found:rxn01208 for OMCDC
Found:rxn02811 for IPPMIa
Found:rxn02789 for IPPMIb
Failure running /Users/jplfaria/Model-SEED-core/lib/ModelSEED/ModelDriver.pl: install_driver(mysql) failed: Can't locate DBD/mysql.pm in @inc (@inc contains: /Users/jplfaria/Model-SEED-core/lib /Users/jplfaria/Model-SEED-core/lib/FigKernelPackages /Users/jplfaria/Model-SEED-core/lib/myRAST /Users/jplfaria/Model-SEED-core/lib/PPO /Users/jplfaria/Model-SEED-core/config/ /opt/local/lib/perl5/site_perl/5.12.3/darwin-multi-2level /opt/local/lib/perl5/site_perl/5.12.3 /opt/local/lib/perl5/vendor_perl/5.12.3/darwin-multi-2level /opt/local/lib/perl5/vendor_perl/5.12.3 /opt/local/lib/perl5/5.12.3/darwin-multi-2level /opt/local/lib/perl5/5.12.3 /opt/local/lib/perl5/site_perl /opt/local/lib/perl5/vendor_perl .) at (eval 1946) line 3.
Perhaps the DBD::mysql perl module hasn't been fully installed,
or perhaps the capitalisation of 'mysql' isn't right.
Available drivers: DBM, ExampleP, File, Gofer, Proxy, SQLite, Sponge.
at /Users/jplfaria/Model-SEED-core/lib/PPO/PPOBackend/MySQL.pm line 61

Just moved to ModelSEED / Model-SEED-core and did the git clone command, and I do not find Model-SEED-core/bin/configureModelSEED.sh file.

in all make files of mfatoolkt/linux
there is a line:
GLPKLIB = $(GLPKDIRECTORY)/src/.libs/

which points to the glpk lib path.

It looks like an error and should maybe /src/../lib/

If a genome id is first used upon import, the server fetches the genome statistics from Sapling/RAST/PubSeed and then spurges out a long list of errors for every role that could be in the local database, possibly because not every role is in the downloaded SQL database (is it?)

In any case, the model import proceeds fine, and doesn't show the error again if the same genome id is used again because the statistics had already been compiled.

To reproduce the error, the corresponding row from the GENOMESTATS table must be removed.

This has just come up, and I can no longer run my model imports. The error I'm getting is a "500" error returned by ClientThing because it's trying to contact:

http://bioseed.mcs.anl.gov/~chenry/FIG/CGI/FBAMODEL_server.cgi

1. I don't know why its doing this, but this cant be integral to model import
2. If you visit this website, you'll see that some perl modules have gone missing...

Hi

I have started from scratch again the installation of the project (Model-SEED-core)

I decided that I will try to work with sqlite so that I will have less problems,

however I still have problems.

Here they are:

First add in the file "Linux and OSX Installation Instructions.txt" the instruction to install DBD-SQLite for those users using sqlite!

1. I did not know what to put in the setting.config "filename=", however as section 4 asks to download the database and you instruct to load it using sqlite3, it can be a good idea to mention the relation between the database and the database file name.

1.section 1 in the "Configuration Instructions.txt" file is not clear enough yet.
The word [optional] appears near parameters which I understand must be filled. this is confusing

1. in section 4 of the same file, the title is "Set up MySQL database backend "
   now in section 2 you point me to this section if I use SQLite,
   It was not clear what you wanted.

Any way I have run the command:sqlite3 Model-SEED-core/data/ModelDB/ModelDB.db < Model-SEED-core/data/ModelDB/ModelDB.sqlite;

and then ran the ms-config

and got the following message.

No user account found with name: public!

I do not know if this is OK or not.

1. and finally in section 5. "Test the Model SEED"
   I ran the command as suggested but I got the following message:
   Function name must be specified as input arguments!

but what is that function name????

For some reason, I stopped getting issue notifications once the Model-SEED-core repository got moved to the ModelSEED organization, so I've fiddled with things and am creating this issue to see if I will start getting notifications again.

Because compilation of the mfatoolkit/Source results in object files that aren't included in the git repository, performing a status will highlight them. I'd add them to .gitignore, but I'm not sure if that's entirely appropriate.

In the installation instructions there is an entry regarding how to test the environment after installation.

.8) Test environment by running Model Driver with no arguments:
a.) Run this script: Model-SEED-core/bin/ModelDriver.(sh/cmd)

When doing so I get:
./ModelDriver.sh
Function name must be specified as input arguments!

I can not learn from this if al is OK or not.
Either say that this is the expected behavior or change the script report

I was following the instructions to install the Model-SEED-core and I got a problem on this step:
c.) Load SQL-lite database from file:
sqlite3 Model-SEED-core/data/ModelDB/ModelDB.db < Model-SEED-core/data/ModelDB/ModelDB.sqlite;

Running the previous command returns only errors like:

mcswl156:~ jplfaria$ sqlite3 ModelSEED/data/ModelDB/ModelDB.db < ModelSEED/data/ModelDB/ModelDB.sqlite
Error: near line 2: table _metainfo already exists
Error: near line 3: PRIMARY KEY must be unique
Error: near line 7: table _references already exists
Error: near line 8: table _objects already exists
Error: near line 9: PRIMARY KEY must be unique
....
Error: near line 10: PRIMARY KEY must be unique
Error: near line 50: PRIMARY KEY must be unique
Error: near line 51: table COMPOUND_GROUPING already exists
Error: near line 52: table SUBSYSTEM already exists
Error: near line 53: table COMMENT already exists
Error: near line 54: table COMPOUND_BIOMASS already exists
Error: near line 55: table MEDIA_COMPOUND already exists
Error: near line 56: table DIAGRAM already exists
Error: near line 57: table SUBSYSTEM_ROLE already exists
...
etc.

José
Error: near line 58: table REACTION_GROUPING already exists

After updating with the latest commit, I got this error when I tried to re-run a model import. I re-ran bin/ms-config with my settings file just in case, and still got the same error.

Stack trace for event: Object class ModelSEED::ModelDB::USERS has no attribute 'check_password'.
Tracer::Confess('Object class ModelSEED::ModelDB::USERS has no attribute 'che...') called at /Users/seaver/Software/ModelSEEDCore/lib/PPO/DBObject.pm line 735
DBObject::AUTOLOAD('ModelSEED::ModelDB::USERS=HASH(0xe1a9e0)', 'bollocks') called at /Users/seaver/Software/ModelSEEDCore/lib/ModelSEED/FIGMODEL.pm line 809
ModelSEED::FIGMODEL::authenticate('ModelSEED::FIGMODEL=HASH(0xdac400)', 'HASH(0xdac5e0)') called at /Users/seaver/Software/ModelSEEDCore/lib/ModelSEED/FIGMODEL.pm line 778
ModelSEED::FIGMODEL::authenticate_user('ModelSEED::FIGMODEL=HASH(0xdac400)', 'seaver', 'bollocks') called at /Users/seaver/Software/ModelSEEDCore/lib/ModelSEED/FIGMODEL.pm line 122
ModelSEED::FIGMODEL::new('ModelSEED::FIGMODEL') called at /Users/seaver/Software/ModelSEEDCore/lib/ModelSEED/ModelDriver.pm line 23
ModelSEED::ModelDriver::new('ModelSEED::ModelDriver') called at /Users/seaver/Software/ModelSEEDCore/lib/ModelSEED/ModelDriver.pl line 26
require /Users/seaver/Software/ModelSEEDCore/lib/ModelSEED/ModelDriver.pl called at /Users/seaver/Software/ModelSEEDCore/config//ModelSEEDbootstrap.pm line 23
main::run() called at -e line 3
Object class ModelSEED::ModelDB::USERS has no attribute 'check_password'.
Failure running /Users/seaver/Software/ModelSEEDCore/lib/ModelSEED/ModelDriver.pl: >>> Object class ModelSEED::ModelDB::USERS has no attribute 'check_password'. at /Users/seaver/Software/ModelSEEDCore/lib/PPO/DBObject.pm line 735

You'll probably want to add a function in configureModelSeed.pl that handles this given the parameters in Settings.txt

mysql> insert into users (firstname,lastname,email,login,password)
values ('Sam','Seaver','[email protected]','seaver','bollocks')

After that, it actually seemed to be running now! The output had some extra 'test1' statements, left over from debugging I presume:

Processing argument:
importmodel?KEGGimport?83333.1?seaver?/Users/seaver/Documents/Projects/Model_Import_Files/?1
Use of uninitialized value in pattern match (m//) at /Users/seaver/Software/Model-SEED-core/lib/PPO/DBObject.pm line 493.
test1 /Users/seaver/Software/Model-SEED-core/models/seaver/KEGGimport.2/0/biochemistry/reaction.txt
test1 /Users/seaver/Software/Model-SEED-core/models/seaver/KEGGimport.2/0/biochemistry/compound.txt
test1 /Users/seaver/Software/Model-SEED-core/models/seaver/KEGGimport.2/0/biochemistry/cpdals.txt
test1 /Users/seaver/Software/Model-SEED-core/models/seaver/KEGGimport.2/0/biochemistry/rxnals.txt
Found:cpd17154
etc. etc.

I'm getting a generic error:

StringDBTable:loadFromFile:could not open table file StringDBTable:get_object:no hash exists for query attribute StringDBTable:get_object:no hash exists for query attribute

It seems that StringDB.cpp in mfatoolkit is getting the wrong filename.

After a bit of debugging, I've found that the command that is run is:

/Users/seaver/Software/ModelSEEDCore/software/mfatoolkit/bin/mfatoolkit resetparameter "MFA input directory" /Users/seaver/Software/ModelSEEDCore/data/ReactionDB/ parameterfile "../Parameters/ProductionMFA.txt" parameterfile "../Parameters/GapFilling.txt" parameterfile "/Users/seaver/Software/ModelSEEDCore/data/ReactionDB/MFAToolkitOutputFiles/2098275/CompleteGapfillingParameters.txt" parameterfile "/Users/seaver/Software/ModelSEEDCore/data/ReactionDB/MFAToolkitOutputFiles/2098275/AddnlFBAParameters.txt" LoadCentralSystem "/Users/seaver/Software/ModelSEEDCore/data/ReactionDB/MFAToolkitOutputFiles/2098275/iJR904.45632.tbl" > "/Users/seaver/Software/ModelSEEDCore/logs/fba-2098275.log"

and its attempting to run this command from

/Users/seaver/Software/ModelSEEDCore

Its my belief that it cannot find some of the parameter files properly.

Right now if a routine called by ModelDriver dies it does not report a "user friendly" error message. We can fix this by catching the error, parsing out the actual error message and displaying it to the user along with the suggestion to submit the detailed error log (the stack trace) to this ticket repository.

I've written an installation script bootstrap.sh.

Called with this syntax:
bootstrap.sh path/to/install

It does the following:

1. Checkout latest version of Model-SEED-core repository
2. Download and install flatfile data from http://bioseed.mcs.anl.gov/~chenry/data.tgz ( in data/ )
3. Downlaod and install ModelDB database (sqlite) from http://bioseed.mcs.anl.gov/~devoid/ModelDB-sqlite.tgz

At that point, you need to edit your configuration, specifically to point ms-config at the GLPK lib dir.

Then you run bin/ms-config load to complete everything.

Ideally we could automatically discover the location of the GLPK directory, then you'd have a working ModelSEED install after running one command. No reading instructions, no additional downloads.

the script "bin/ms-load-mysql" is not available (as said in the wiki)

As per a previous recommendation by Chris, I'm using genome id 83333.1, which is E. coli of course, the model organism. However, I'm using it for generic databases, and also, users may import models for which there is no genome installed in SEED (and therefore no genome id).

Running modelimport on a generic id like 00000.0 doesn't work, so this should probably be added, either use a generic string, or even allow a model to be imported with no genome id.

In the BIO environment, it didn't print these files out locally, only in a pre-specified location:

/vol/model-dev/MODEL_DEV_DB/MSModelFiles/iJR904.tbl
/vol/model-dev/MODEL_DEV_DB/MSModelFiles/bio00320.txt

In the local environment, it didn't recognize my unique SEED ID, even if I "log in" so I had to append it to the model name.

Finally, it's expecting the biochemistry data to be in the database rather than in the flat files, because I get this error:

"Cannot obtain ppo data for reaction"

There probably should be a short example, perhaps one reaction with several compounds, in two files, to test the model import with. In addition, there I suggest some accompanying documentation that lists the required headers and formats (even I am still not sure which columns are optional).

When I run the KEGG import, I still get this error message for a limited number of reactions:

"Empty products or products string."

The first line of Parameters/Defaults.txt:

database spec file|../../Input/Singlefile.txt|

refers to a file that doesn't exist, what are the contents supposed to be?

Need to be able to switch a core between different configurations. Something like:

$ ms-config load config/foo.txt
$ ms-config unload
$ ms-config load config/bar.txt
$ ms-config reload config/foo.txt
$ ms-config status

I get these two errors/warnings consistenty when using fbafva and testmodelgrowth:

~/Desktop/ModelSEED/Model-SEED-core> bin/fbafva -model "iYusMp.149"

Processing argument: fbafva
[PPOBackend.pm 246]Database error: DBD::mysql::st execute failed: Unknown column 'defaultStudyMedi
a' in 'field list' at /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/lib/PPO/PPOBackend.pm line 2
29.
mkdir: /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/data/ReactionDB/MFAToolkitOutputFiles/20982
67/MFAOutput: File exists

~/Desktop/ModelSEED/Model-SEED-core/bin> ModelDriver classifyreactions?iJR904.3?Complete?0?0

Processing argument: classifyreactions?iJR904.3?Complete?0?0
Use of uninitialized value in pattern match (m//) at /Users/dejongh/Desktop/ModelSEED/Model-SEED-co
re/lib/PPO/DBObject.pm line 495.
Now processing model: iJR904.3
/Users/dejongh/Desktop/ModelSEED/Model-SEED-core/software/mfatoolkit/bin/mfatoolkit resetparameter
"MFA input directory" /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/data/ReactionDB/ parameterfi
le "../Parameters/ProductionMFA.txt" parameterfile "/Users/dejongh/Desktop/ModelSEED/Model-SEED-cor
e/data/ReactionDB/MFAToolkitOutputFiles/2098263/FluxVariabilityAnalysis.txt" parameterfile "/Users/
dejongh/Desktop/ModelSEED/Model-SEED-core/data/ReactionDB/MFAToolkitOutputFiles/2098263/AddnlFBAPar
ameters.txt" LoadCentralSystem "/Users/dejongh/Desktop/ModelSEED/Model-SEED-core/data/ReactionDB/MF
AToolkitOutputFiles/2098263/iJR904.3.tbl" > "/Users/dejongh/Desktop/ModelSEED/Model-SEED-core/logs/
fba-2098263.log"
StringDBTable:loadFromFile:could not open table file
StringDBTable:loadFromFile:could not open table file
mkdir: /scratch: No such file or directory
cp: /scratch/2098263: No such file or directory
FIGMODELfba:parseTightBounds:could not find tight bound results file
SUCCESS:
FAIL:iJR904.3;

Hi

Just run the command listed in wiki :./importmodel -name "iJR904" -genome "83333.1" after login
and got
Processing argument: importmodel
Found:cpd00201 for 10fthf
Found:cpd11423 for 12dgr_EC
....
Found:rxn02811 for IPPMIa
Found:rxn02789 for IPPMIb
Whoops! We encountered an unrecoverable error.

Creating new object failed.

A complete log of the problem has been printed to:

/home/raphyz/Desktop/SHARE/phd/Model-SEED-core/.errors/error-ei5dc

this log file content is:

ModelDriver encountered an unrecoverable error:

""Creating new object failed."" at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/ModelSEED/FIGMODEL.pm line 133
ModelSEED::FIGMODEL::FIGMODELERROR('Creating new object failed.') called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/PPO/DBObject.pm line 370
DBObject::set_attributes('ModelSEED::ModelDB::REACTION_MODEL=HASH(0x1e389008)', 'HASH(0x99a94b0)') called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/PPO/DBObject.pm line 239
DBObject::create('ModelSEED::ModelDB::REACTION_MODEL=HASH(0x1e389008)', 'HASH(0x99a94b0)') called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/ModelSEED/FIGMODEL/FIGMODELdatabase.pm line 232
ModelSEED::FIGMODEL::FIGMODELdatabase::create_object('ModelSEED::FIGMODEL::FIGMODELdatabase=HASH(0x4b9c900)', 'rxnmdl', 'HASH(0x99a94b0)') called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/ModelSEED/FIGMODEL.pm line 2830
ModelSEED::FIGMODEL::import_model('ModelSEED::FIGMODEL=HASH(0x2e6ed10)', 'HASH(0x3009a28)') called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/ModelSEED/ModelDriver.pm line 3959
ModelSEED::ModelDriver::importmodel('ModelSEED::ModelDriver=HASH(0x2e6f058)', 'importmodel', 'HASH(0x300d498)') called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/ModelSEED/ModelDriver.pl line 121
main::ANON() called at /usr/lib64/perl5/vendor_perl/5.12.3/Try/Tiny.pm line 76
eval {...} called at /usr/lib64/perl5/vendor_perl/5.12.3/Try/Tiny.pm line 67
Try::Tiny::try('CODE(0x2e6f0b8)', 'Try::Tiny::Catch=REF(0x2e66b00)') called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/ModelSEED/ModelDriver.pl line 124
require /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/lib/ModelSEED/ModelDriver.pl called at /home/raphyz/Desktop/SHARE/phd/Model-SEED-core/config//ModelSEEDbootstrap.pm line 27
main::run() called at -e line 3

@samseaver
I was trying to run the gap-filling on the latest update, but it had this error:

ModelDriver encountered an unrecoverable error:

Can't call method "serialize" on an undefined value at
/vol/rast-bcr/2010-1124/linux-rhel5-x86_64/lib/perl5/site_perl/5.12.2/CHI/CacheObject.pm
line 57

I'm hoping this is easy to fix :-)

BTW, I got File::NSFLock and CHI::?? to install...

dejongh@biologin-4:~/Model-SEED-core$ /vol/model-prod/env/config.sh: line 2: /home/dejongh/Model-SEED-core/Model-SEED-core/bin/ms-config: No such file or directory

[2]+ Exit 127 /vol/model-prod/env/install.sh /home/dejongh/Model-SEED-core (wd: ~)

EM -c ../Source/driver.cpp; mv .o ../Source
../Source/driver.cpp:22:24: error: MFAToolkit.h: No such file or directory
../Source/driver.cpp:124:3: warning: no newline at end of file
../Source/driver.cpp: In function 'int main(int, char_)':
../Source/driver.cpp:27: error: 'SetProgramPath' was not declared in this scope
../Source/driver.cpp:28: error: 'SetParameter' was not declared in this scope
../Source/driver.cpp:31: error: 'STATIC_INPUT_FILE' was not declared in this scope
../Source/driver.cpp:31: error: 'SetInputParametersFile' was not declared in this scope
../Source/driver.cpp:32: error: 'LoadParameters' was not declared in this scope
../Source/driver.cpp:32: error: 'FAIL' was not declared in this scope
../Source/driver.cpp:35: error: 'ClearParameterDependance' was not declared in this scope
../Source/driver.cpp:36: error: 'Initialize' was not declared in this scope
../Source/driver.cpp:36: error: 'FAIL' was not declared in this scope
../Source/driver.cpp:39: error: 'LoadFIGMODELParameters' was not declared in this scope
../Source/driver.cpp:41: error: 'DriverOptions' was not declared in this scope
../Source/driver.cpp:43: error: 'vector' was not declared in this scope
../Source/driver.cpp:43: error: 'string' was not declared in this scope
../Source/driver.cpp:43: error: 'Arguments' was not declared in this scope
../Source/driver.cpp:45: error: expected ;' before 'Buff' ../Source/driver.cpp:46: error: 'Buff' was not declared in this scope ../Source/driver.cpp:48: error: 'CommandlineInterface' was not declared in this scope ../Source/driver.cpp:51: error: 'Cleanup' was not declared in this scope ../Source/driver.cpp:54: error: 'string' was not declared in this scope ../Source/driver.cpp:54: error: expected;' before 'OutputPath'
../Source/driver.cpp:55: error: 'OutputPath' was not declared in this scope
../Source/driver.cpp:56: error: 'GetParameter' was not declared in this scope
../Source/driver.cpp:58: error: 'GetDatabaseDirectory' was not declared in this scope
../Source/driver.cpp:58: error: 'system' was not declared in this scope
../Source/driver.cpp:60: error: 'GetDatabaseDirectory' was not declared in this scope
../Source/driver.cpp:60: error: 'system' was not declared in this scope
../Source/driver.cpp: In function 'void DriverOptions()':
../Source/driver.cpp:69: error: 'cout' was not declared in this scope
../Source/driver.cpp:69: error: 'endl' was not declared in this scope
../Source/driver.cpp:78: error: 'AskNum' was not declared in this scope
../Source/driver.cpp:82: error: 'LoadDatabaseFile' was not declared in this scope
../Source/driver.cpp:86: error: 'ProcessDatabase' was not declared in this scope
../Source/driver.cpp:93: error: 'ProcessWebInterfaceModels' was not declared in this scope
mv: cannot stat `_.o': No such file or directory
make: *** [../Source/driver.o] Error 1
No user account found with name: !

Could not find new biomass reaction bio00001 at /Users/best/ModelSEED/lib/ModelSEED/FIGMODEL.pm line 137
ModelSEED::FIGMODEL::FIGMODELWARNING('Could not find new biomass reaction bio00001') called at /Users/best/ModelSEED/lib/ModelSEED/FIGMODEL/FIGMODELmodel.pm line 1587
ModelSEED::FIGMODEL::FIGMODELmodel::biomassReaction('ModelSEED::FIGMODEL::FIGMODELmodel=HASH(0xd31ec0)', 'bio00001') called at /Users/best/ModelSEED/lib/ModelSEED/FIGMODEL.pm line 2648
ModelSEED::FIGMODEL::import_model('ModelSEED::FIGMODEL=HASH(0xcf2170)', 'HASH(0x815550)') called at /Users/best/ModelSEED/lib/ModelSEED/ModelDriver.pm line 4043
ModelSEED::ModelDriver::importmodel('ModelSEED::ModelDriver=HASH(0x8153c0)', 'importmodel', 'HASH(0xd47e00)') called at /Users/best/ModelSEED/lib/ModelSEED/ModelDriver.pl line 61
require /Users/best/ModelSEED/lib/ModelSEED/ModelDriver.pl called at /Users/best/ModelSEED/config//ModelSEEDbootstrap.pm line 24
main::run() called at -e line 3
Failure running /Users/best/ModelSEED/lib/ModelSEED/ModelDriver.pl: Can't call method "public" on an undefined value at /Users/best/ModelSEED/lib/ModelSEED/FIGMODEL/FIGMODELmodel.pm line 659.

FBAMODEL.t is currently consuming ~4G of memory and has been running for 5 minutes. We need to try to get this down to an acceptable level. (< 1 minute for all tests)

[PPOBackend.pm 244]Database error: DBD::SQLite::db prepare failed: no such column: defaultStudyMedia at /Users/jplfaria/Model-SEED-core/lib/PPO/PPOBackend.pm line 226.

Found:rxn00191 for ALATA_L
Found:rxn00283 for ALAR !!!!!!!!!!!!!!!!!!!!!!!!!!!
Found:rxn00342 for ASNN
Found:rxn00340 for ASNS2
Found:rxn00416 for ASNS1
Found:rxn00347 for ASPT
Found:rxn00260 for ASPTA
Found:rxn00904 for VPAMT
Found:rxn08291 for DAAD
Found:rxn00283 for ALARi !!!!!!!!!!!!!!!!!!!!!!!!!
Failure running /Users/jplfaria/Model-SEED-core/lib/ModelSEED/ModelDriver.pl: Creating new object failed. at /Users/jplfaria/Model-SEED-core/lib/ModelSEED/FIGMODEL.pm line 132
ModelSEED::FIGMODEL::FIGMODELERROR('Creating new object failed.') called at /Users/jplfaria/Model-SEED-core/lib/PPO/DBObject.pm line 370
DBObject::set_attributes('ModelSEED::ModelDB::REACTION_MODEL=HASH(0x7fa154a09390)', 'HASH(0x7fa13c5057d8)') called at /Users/jplfaria/Model-SEED-core/lib/PPO/DBObject.pm line 239
DBObject::create('ModelSEED::ModelDB::REACTION_MODEL=HASH(0x7fa154a09390)', 'HASH(0x7fa13c5057d8)') called at /Users/jplfaria/Model-SEED-core/lib/ModelSEED/FIGMODEL/FIGMODELdatabase.pm line 232
ModelSEED::FIGMODEL::FIGMODELdatabase::create_object('ModelSEED::FIGMODEL::FIGMODELdatabase=HASH(0x7fa13d844150)', 'rxnmdl', 'HASH(0x7fa13c5057d8)') called at /Users/jplfaria/Model-SEED-core/lib/ModelSEED/FIGMODEL.pm line 2840
ModelSEED::FIGMODEL::import_model('ModelSEED::FIGMODEL=HASH(0x7fa13c068ff8)', 'HASH(0x7fa13c0e7368)') called at /Users/jplfaria/Model-SEED-core/lib/ModelSEED/ModelDriver.pm line 4043
ModelSEED::ModelDriver::importmodel('ModelSEED::ModelDriver=HASH(0x7fa13a02f168)', 'importmodel', 'HASH(0x7fa13a30d568)') called at /Users/jplfaria/Model-SEED-core/lib/ModelSEED/ModelDriver.pl line 61
require /Users/jplfaria/Model-SEED-core/lib/ModelSEED/ModelDriver.pl called at /Users/jplfaria/Model-SEED-core/config//ModelSEEDbootstrap.pm line 24
main::run() called at -e line 3

I scp'ed my iYusMp.149 model from bio* to my desktop, then ran

bin/loadmodelfromfile -name "iYusMp.149" -genome "272634.1"

Got the following error log (note that the id used below has an extra ".149" on it)

ModelDriver encountered an unrecoverable error:

""Could not find model specification file: /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/data/MS
ModelFiles/iYusMp.149.149.tbl!"" at /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/lib/ModelSEED/
FIGMODEL.pm line 133
ModelSEED::FIGMODEL::FIGMODELERROR('Could not find model specification file: /Users/dejongh
/Deskt...') called at /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/lib/ModelSEED/FIGMODEL.pm li
ne 2545
ModelSEED::FIGMODEL::import_model_file('ModelSEED::FIGMODEL=HASH(0x7fbcac03fe50)', 'HASH(0x
7fbcae21a6b0)') called at /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/lib/ModelSEED/ModelDrive
r.pm line 4513
ModelSEED::ModelDriver::loadmodelfromfile('ModelSEED::ModelDriver=HASH(0x7fbcac040198)', 'l
oadmodelfromfile', 'HASH(0x7fbcae21a5c0)') called at /Users/dejongh/Desktop/ModelSEED/Model-SEED-co
re/lib/ModelSEED/ModelDriver.pl line 121
main::ANON() called at /System/Library/Perl/Extras/5.12/Try/Tiny.pm line 76
eval {...} called at /System/Library/Perl/Extras/5.12/Try/Tiny.pm line 67
Try::Tiny::try('CODE(0x7fbcac0401f8)', 'Try::Tiny::Catch=REF(0x7fbcae1a41d0)') called at /U
sers/dejongh/Desktop/ModelSEED/Model-SEED-core/lib/ModelSEED/ModelDriver.pl line 124
require /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/lib/ModelSEED/ModelDriver.pl calle
d at /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/config//ModelSEEDbootstrap.pm line 26
main::run() called at -e line 3

I reran using

bin/loadmodelfromfile -name "iYusMp" -genome "272634.1"

and didn't get any errors log, but did get the following message:

Processing argument: loadmodelfromfile
Adding new model iYusMp.149 for mdejongh into database
[PPOBackend.pm 246]Database error: DBD::mysql::st execute failed: Unknown column 'defaultStudyMedi
a' in 'field list' at /Users/dejongh/Desktop/ModelSEED/Model-SEED-core/lib/PPO/PPOBackend.pm line 2
29.
Successfully imported iYusMp into Model SEED as iYusMp.149!

The import gives this error when it starts:

Use of uninitialized value in pattern match (m//) at /Users/seaver/Software/Model-SEED-core/lib/PPO/DBObject.pm line 493.

The model import proceeds and completes just fine, but this error is posted here in case it has the potential to be a bug.

I've not tried this in the BIO environment, but locally:

1. The -biomass parameter doesn't work, it keeps looking in the same location (./data/MSModelFiles)

2. The -filename parameter works, but if it is not used, then the filename that it looks for must have the user index in the filename, ie:

data/MSModelFiles/iJR904.45632.tbl1

instead of

data/MSModelFiles/iJR904.tbl

3. The overwrite function works for sake of documentation, but if the script fails the first time (ie can't find biomass file) it doesn't roll back the inserted data, so the next time you run it, even though it "failed" you still have to overwrite. This can be confusing.

NOTE FROM SAM:

I made some progress on this. The config directory didn't exist, so I made the directory, your instructions implied that it should already be there in the git respository. The Settings.txt file was found fine.

configureModelSEED.sh worked, and successfully compiled with glpk.
There was a warning in the compilation, which I've pasted here for your reference:

../Source/MFAProblem.cpp: In member function ‘int MFAProblem::ModifyInputConstraints(ConstraintsToModify*, Data*)’:
../Source/MFAProblem.cpp:1828: warning: NULL used in arithmetic
../Source/MFAProblem.cpp:1832: warning: NULL used in arithmetic

I found out when I tried running CreateDBScheme.sh that I didn't have the perl module DBD::mysql installed, so you should put that in the list of perl modules, both DBI, and DBD::mysql. I think you should move the instructions for mysql in front of the instructions for configureModelSEED.sh, because one should create the user account in for the local modeldb database before setting it in the Settings.txt file.

If anyone has trouble getting DBD::mysql to work, remind them that they need the MySQL development headers/libraries as well as the standard installation.

Running CreateDBScheme.sh worked after that, and created a ModelDB2 database in my mysql server. However, when I tried to load in the mysql dump:

seaver$ mysql -u ModelSEED -p ModelDB2 < ModelSEEDDB.sql Enter password:
ERROR 1064 (42000) at line 161: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'USING BTREE
) ENGINE=MyISAM AUTO_INCREMENT=86873 DEFAULT CHARSET=latin1' at line 33

Now, I realize that I'm going to upgrade my mysql server, which is version 5.0, in case there's a change in the SQL format that's relevant for the server version, but you should probably see if you can create a dump that works for most versions of MySQL, as others may not have this flexibility. This might not be the problem, but I'm not sure what else it could be.

On a related note, CreateDBScheme.sh created ModelDB2, but the database dump creates and uses ModelDB, so the ModelDB2 database in fact doesn't get used, which kind of makes CreateDBScheme.sh redundant.

FOLLOW UP NOTE FROM SAM:

OK, I found this bug report: http://bugs.mysql.com/bug.php?id=25162

It suggests running mysqldump with the --compatible option:

--compatible=name   Change the dump to be compatible with a given mode. By
                      default tables are dumped in a format optimized for
                      MySQL. Legal modes are: ansi, mysql323, mysql40,
                      postgresql, oracle, mssql, db2, maxdb, no_key_options,
                      no_table_options, no_field_options. One can use several
                      modes separated by commas. Note: Requires MySQL server
                      version 4.1.0 or higher. This option is ignored with
                      earlier server versions.

Can you try it as:

mysqldump --compatible=mysql40

and send me the new sql dump?

MORE FOLLOW UP FROM SAM:

Actually, I went through that sql dump, and deleted all 20 instances of "USING BTREE", and the dump worked fine after that.