Reference

TFBS::DB::JASPAR2(3pm)

interface to MySQL relational database of pattern matrices

NAME

TFBS::DB::JASPAR2 - interface to MySQL relational database of pattern matrices

SYNOPSIS

•

creating a database object by connecting to the existing JASPAR2-type database

    my $db = TFBS::DB::JASPAR2->connect("dbi:mysql:JASPAR2:myhost",
                                        "myusername",
                                        "mypassword");
•

retrieving a TFBS::Matrix::* object from the database

    # retrieving a PFM by ID
    my $pfm = $db->get_Matrix_by_ID('M0079','PFM');
 
    #retrieving a PWM by name
    my $pwm = $db->get_Matrix_by_name('NF-kappaB', 'PWM');
•

retrieving a set of matrices as a TFBS::MatrixSet object according to various criteria

    # retrieving a set of PWMs from a list of IDs:
    my @IDlist = ('M0019', 'M0045', 'M0073', 'M0101');
    my $matrixset = $db->get_MatrixSet(-IDs => \@IDlist,
                                       -matrixtype => "PWM");
 
    # retrieving a set of ICMs from a list of names:
    my @namelist = ('p50', 'p53', 'HNF-1'. 'GATA-1', 'GATA-2', 'GATA-3');
    my $matrixset = $db->get_MatrixSet(-names => \@namelist,
                                       -matrixtype => "ICM");
 
    # retrieving a set of all PFMs in the database
    # derived from human genes:
    my $matrixset = $db->get_MatrixSet(-species => ['Homo sapiens'],
                                       -matrixtype => "PFM");
•

creating a new JASPAR2-type database named MYJASPAR2:

    my $db = TFBS::DB::JASPAR2->create("dbi:mysql:MYJASPAR2:myhost",
                                       "myusername",
                                       "mypassword");
•

storing a matrix in the database (currently only PFMs):

    #let $pfm is a TFBS::Matrix::PFM object
    $db->store_Matrix($pfm);

DESCRIPTION

TFBS::DB::JASPAR2 is a read/write database interface module that retrieves and stores TFBS::Matrix::* and TFBS::MatrixSet objects in a relational database.

JASPAR2 DATA MODEL

JASPAR2 is working name for a relational database model used for storing transcriptional factor pattern matrices in a MySQL database. It was initially designed to store matrices for the JASPAR database of high quality eukaryotic transcription factor specificity profiles by Albin Sandelin and Wyeth W. Wasserman. Besides the profile matrix itself, this data model stores profile ID (unique), name, structural class, basic taxonomic and bibliographic information as well as some additional optional tags.

Due to its data model, which precedeed the design of the module, TFBS::DB::JASPAR2 cannot store arbitrary tags for a matrix.

The supported tags are
'acc' # (accession number;
# originally for transcription factor protein seq)
'seqdb' # sequence database where 'acc' comes from
'medline' # PubMed ID
'species' # Species name
'sysgroup'
'total_ic' # total information content - redundant, present
# for historical "medline" => ($self->_get_medline($ID) or ""),
"species" => ($self->_get_species($ID) or ""),
"sysgroup"=> ($self->_get_sysgroup($ID) or ""),
"type" => ($self->_get_type($ID) or ""),
"seqdb" => ($self->_get_seqdb($ID) or ""),
"acc" => ($self->_get_acc($ID) or ""),
"total_ic"=

----------------------- ADVANCED ---------------------------------

For the developers and the curious, here is the JASPAR2 data model:

       CREATE TABLE matrix_data (
         ID varchar(16) DEFAULT '' NOT NULL,
         pos_ID varchar(24) DEFAULT '' NOT NULL,
         base enum('A','C','G','T'),
         position tinyint(3) unsigned,
         raw int(3) unsigned,
         info float(7,5) unsigned, -- calculated
         pwm float(7,5) unsigned,  -- calculated
         normalized float(7,5) unsigned,
         PRIMARY KEY (pos_ID),
         KEY id_index (ID)
       );
       CREATE TABLE matrix_info (
         ID varchar(16) DEFAULT '' NOT NULL,
         name varchar(15) DEFAULT '' NOT NULL,
         type varchar(8) DEFAULT '' NOT NULL,
         class varchar(20),
         phylum varchar (32),          -- maps to 'sysgroup' tag
         litt varchar(40),             -- not used by this module
         medline int(12),
         information varchar(20),      -- not used by this module
         iterations varchar(6),
         width int(2),                 -- calculated
         consensus varchar(25),        -- calculated
         IC float(6,4),                -- maps to 'total_ic' tag
         sites int(3) unsigned,        -- not used by this module
         PRIMARY KEY (ID)
       )
       CREATE TABLE matrix_seqs (
         ID varchar(16) DEFAULT '' NOT NULL,
         internal varchar(8) DEFAULT '' NOT NULL,
         seq_db varchar(15) NOT NULL,
         seq varchar(10) NOT NULL,
         PRIMARY KEY (ID, seq_db, seq)
       )
       CREATE TABLE matrix_species (
         ID varchar(16) DEFAULT '' NOT NULL,
         internal varchar(8) DEFAULT '' NOT NULL,
         species varchar(24) NOT NULL,
         PRIMARY KEY (ID, species)
       )

It is our best intention to hide the details of this data model, which we are using on a daily basis in our work, from most TFBS users, simply because for historical reasons some table column names are confusing at best. Most users should only know the methods to store the data and which tags are supported.

-------------------------------------------------------------------------

FEEDBACK

Please send bug reports and other comments to the author.

AUTHOR - Boris Lenhard

Boris Lenhard <[email protected]>

APPENDIX

The rest of the documentation details each of the object methods. Internal methods are preceded with an underscore.

new

 Title   : new
 Usage   : DEPRECATED - for backward compatibility only
           Use connect() or create() instead

connect

 Title   : connect
 Usage   : my $db =
            TFBS::DB::JASPAR2->connect("dbi:mysql:DATABASENAME:HOSTNAME",
                                        "USERNAME",
                                        "PASSWORD");
 Function: connects to the existing JASPAR2-type database and
           returns a database object that interfaces the database
 Returns : a TFBS::DB::JASPAR2 object
 Args    : a standard database connection triplet
           ("dbi:mysql:DATABASENAME:HOSTNAME",  "USERNAME", "PASSWORD")
           In place of DATABASENAME, HOSTNAME, USERNAME and PASSWORD,
           use the actual values. PASSWORD and USERNAME might be
           optional, depending on the user access permissions for
           the database server.

create

 Title   : create
 Usage   : my $newdb =
            TFBS::DB::JASPAR2->create("dbi:mysql:NEWDATABASENAME:HOSTNAME",
                                      "USERNAME",
                                      "PASSWORD");
 Function: connects to the database server, creates a new JASPAR2-type database and returns a database
           object that interfaces the database
 Returns : a TFBS::DB::JASPAR2 object
 Args    : a standard database connection triplet
            ("dbi:mysql:NEWDATABASENAME:HOSTNAME",  "USERNAME", "PASSWORD")
           In place of NEWDATABASENAME, HOSTNAME, USERNAME and
           PASSWORD use the actual values. PASSWORD and USERNAME
           might be optional, depending on the users access permissions
           for the database server.

dbh

 Title   : dbh
 Usage   : my $dbh = $db->dbh();
           $dbh->do("UPDATE matrix_data SET name='ADD1' WHERE NAME='SREBP2'");
 Function: returns the DBI database handle of the MySQL database
           interfaced by $db; THIS IS USED FOR WRITING NEW METHODS
           FOR DIRECT RELATIONAL DATABASE MANIPULATION - if you
           have write access AND do not know what you are doing,
           you can severely  corrupt the data
           For documentation about database handle methods, see L<DBI>
 Returns : the database (DBI) handle of the MySQL JASPAR2-type
           relational database associated with the TFBS::DB::JASPAR2
           object
 Args    : none

get_Matrix_by_ID

 Title   : get_Matrix_by_ID
 Usage   : my $pfm = $db->get_Matrix_by_ID('M00034', 'PFM');
 Function: fetches matrix data under the given ID from the
           database and returns a TFBS::Matrix::* object
 Returns : a TFBS::Matrix::* object; the exact type of the
           object depending on the second argument (allowed
           values are 'PFM', 'ICM', and 'PWM'); returns undef if
           matrix with the given ID is not found
 Args    : (Matrix_ID, Matrix_type)
           Matrix_ID is a string; Matrix_type is one of the
           following: 'PFM' (raw position frequency matrix),
           'ICM' (information content matrix) or 'PWM' (position
           weight matrix)
           If Matrix_type is omitted, a PWM is retrieved by default.

get_Matrix_by_name

 Title   : get_Matrix_by_name
 Usage   : my $pfm = $db->get_Matrix_by_name('HNF-1', 'PWM');
 Function: fetches matrix data under the given name from the
           database and returns a TFBS::Matrix::* object
 Returns : a TFBS::Matrix::* object; the exact type of the object
           depending on the second argument (allowed values are
           'PFM', 'ICM', and 'PWM')
 Args    : (Matrix_name, Matrix_type)
           Matrix_name is a string; Matrix_type is one of the
           following:
           'PFM' (raw position frequency matrix),
           'ICM' (information content matrix) or
           'PWM' (position weight matrix)
           If Matrix_type is omitted, a PWM is retrieved by default.
 Warning : According to the current JASPAR2 data model, name is
           not necessarily a unique identifier. In the case where
           there are several matrices with the same name in the
           database, the function fetches the first one and prints
           a warning on STDERR. You have been warned.

get_MatrixSet

 Title   : get_MatrixSet
 Usage   : my $matrixset = $db->get_MatrixSet(%args);
 Function: fetches matrix data under for all matrices in the database
           matching criteria defined by the named arguments
           and returns a TFBS::MatrixSet object
 Returns : a TFBS::MatrixSet object
 Args    : This method accepts named arguments:
           -IDs        # a reference to an array of IDs (strings)
           -names      # a reference to an array of
                       #  transcription factor names (string)
           -classes    # a reference to an array of
                       #  structural class names (strings)
           -species    # a reference to an array of
                       #   Latin species names (strings)
           -sysgroups  # a reference to an array of
                       #  higher taxonomic categories (strings)
           -matrixtype # a string, 'PFM', 'ICM' or 'PWM'
           -min_ic     # float, minimum total information content
                       #   of the matrix

The five arguments that expect list references are used in database query formulation: elements within lists are combined with 'OR' operators, and the lists of different types with 'AND'. For example,

    my $matrixset = $db->(-classes => ['TRP_CLUSTER', 'FORKHEAD'],
                          -species => ['Homo sapiens', 'Mus musculus'],
                          -matrixtype => 'PWM');

gives a set of PWMs whose (structural clas is 'TRP_CLUSTER' OR 'FORKHEAD') AND (the species they are derived from is 'Homo sapiens' OR 'Mus musculus').

The -min_ic filter is applied after the query in the sense that the matrices profiles with total information content less than specified are not included in the set.

store_Matrix

 Title   : store_Matrix
 Usage   : $db->store_Matrix($pfm);
 Function: Stores the contents of a TFBS::Matrix::DB object in the database
 Returns : 0 on success; $@ contents on failure
           (this is too C-like and may change in future versions)
 Args    : (PFM_object)
           A TFBS::Matrix::PFM object
 Comment : this is an experimental method that is not 100% bulletproof;
           use at your own risk

store_MatrixSet

 Title   : store_MatrixSet
 Usage   : $db->store_Matrix($matrixset);
 Function: Stores the TFBS::DB::PFM object that are part of a
           TFBS::MatrixSet object into the database
 Returns : 0 on success; $@ contents on failure
           (this is too C-like and may change in future versions)
 Args    : (MatrixSet_object)
           A TFBS::MatrixSet object
 Comment : THIS METHOD IS NOT YET IMPLEMENTED

delete_Matrix_having_ID

 Title   : delete_Matrix_having_ID
 Usage   : $db->delete_Matrix_with_ID('M00045');
 Function: Deletes the matrix having the given ID from the database
 Returns : 0 on success; $@ contents on failure
           (this is too C-like and may change in future versions)
 Args    : (ID)
           A string
 Comment : Yeah, yeah, 'delete_Matrix_having_ID' is a stupid name
           for a method, but at least it should be obviuos what it does.

History