This set of tools defines the schema of a (PostgreSQL) database, intended for storing the results of vast sets of simulations with various incarnations of the Guyton model of whole-body physiology (such as this version).

Prerequisite software:

• Python
• psycopg2
• PostgreSQL

The owner of the database should create an account (a role) in PostgreSQL. The name of the role should correspond to the user name of the owner's account on the host computer. In the following example, the role rma is created:

# sudo -u postgres psql
template1=> CREATE ROLE rma WITH CREATEDB CREATEROLE LOGIN;
template1=> \q

If the first command (su postgres) asks for a password, and trusted connections are enabled (see the PostgreSQL documentation for pg_hba.conf), the following example can be used instead:

# psql -U postgres
template1=> CREATE ROLE rma WITH CREATEDB CREATEROLE LOGIN;
template1=> \q

Once you have created a role, you can access PostgreSQL through the interactive terminal psql:

# psql template1

Here you can enter any SQL commands (if you have the necessary privileges) and psql commands. Type help for more information, or \q to quit.

The file psql_conf.sh contains the configuration settings for the various scripts that create and manipulate the pg_guyton database. Check that these settings are correct before continuing.

The script create_db.sh will create the database according to the schema in schema.sql. Once you have a role that has been assigned the LOGIN and CREATEDB privileges, you can execute the following script:

# ./create_db.sh

The previous step created the database as per the schema, but did not populate any of the tables. To add the necessary parameters and variables to the Guyton database, execute the following script:

# ./create_metadata.py metadata.txt

Each person that needs direct access to the database needs a role whose name corresponds to their user name on the host computer. In the following example, a role is created for the user abc:

# ./create_role.sh abc

Once each user has been assigned a role and the database has been configured, simulation results can be imported using the following script:

# ./import_results.py -t "some tag name" results.txt

Any number of tags can be associated with each of the imported simulations, by providing multiple -t "tag name" arguments. Any tags that don't already exist in the database will be automatically created.

NOTE: The import_results.py script currently only supports perturbation experiments performed with the Guyton model (1992 version).

A separate database of virtual individuals (i.e., pairs of parameter values and the resulting steady-state variables) can be created using the script csv_for_import.pl and psql:

# csv_for_import.pl -v -f data1.txt -f data2.txt
# cat individuals.csv | psql --dbname DATABASE \
  -c "COPY individual (id, perturbed) FROM STDIN WITH CSV"
# cat indiv_params.csv | psql --dbname DATABASE \
  -c "COPY indiv_param (individual, parameter, value) FROM STDIN WITH CSV"
# cat indiv_vars.csv | psql --dbname DATABASE \
  -c "COPY indiv_var (individual, variable, value) FROM STDIN WITH CSV"

On Mac OS X, assuming that PostgreSQL has been installed via macports, the server is started and stopped using the following commands:

# PGDAEMON=`/Library/LaunchDaemons/org.macports.postgresqlXX-server.plist`
# sudo launchctl load -w $PGDAEMON
# sudo launchctl unload -w $PGDAEMON

Remember to replace the XX in $PGDAEMON with the appropriate version number (e.g., '84' for PostgreSQL 8.4 and '91' for PostgreSQL 9.1).