Omics Visualizer version 1.3.0
Last update: 2020-03-31
- Connect a table with a network
- Disconnect
- Apply a filter
- Delete a filter
- Import node table
- Load
- Retrieve a STRING network
- Delete a table
- List tables
- Change the current table
- Pie chart with continuous mapping
- Pie chart with discrete mapping
- Donut chart with continuous mapping
- Donut chart with discrete mapping
- Remove pie chart
- Remove donut chart
ov connect
Connect the current table with the current network
-
mappingColNet
(required) StringKey column from the Network node table used to map the network with the table. The column should be a shared column.
-
mappingColTable
(required) StringKey column from the Omics Visualizer table used to map the table with the network.
ov connect mappingColNet="gene name" mappingColTable="name"
This will link the active OV table with the current network. The OV table must have a column named name
and the node table from the network must have a column named gene name
. Those two columns are used to link the two tables.
List of commands - List of Manage tables commands
ov connect show
Show the connect window of the current table
List of commands - List of Access GUI dialogs commands
ov disconnect
Disconnect the current table and the current network if they are already connected
List of commands - List of Manage tables commands
ov filter
Filters the row of an Omics Visualizer table
-
filter
(required) StringThe filter is defined by the following non-contextual grammar:
filter = and | or | criteria
and = {filter_list}
or = [filter_list]
filter_list = filter,filter_list | filter
criteria = (colName,operator,value)where 'colName' is the name of the column (commas in colName should be escaped by being preceeded a backslash)
'operator' is the name of the operation applied, the list of operators is available with the commandov filter list operators
'value' if necessary, is the value to compare with. Careful with the regex and escaped characters. -
tableName
(optional) StringOV table name.
By default the active table is used, if you want to use another one you can specify its name here.
ov filter filter="{(p-value,LOWER_EQUALS,0.05),[(ontology,CONTAINS,GO),(ontology,NULL,)]}"
This will select the rows where the p-value is lower or equals to 0.05 and the ontology is either from the Gene Ontology (contains "GO") or unknown (the value is null).
List of commands - List of Manage tables commands
ov filter list operators
List the available operators
The list of current available operators is:
EQUALS
(String, boolean, numbers) the table value should match exactly the reference valueNOT_EQUALS
(String, boolean, numbers) the table value should not match the reference valueCONTAINS
(String) a part of the table value should match exactly the reference valueNOT_CONTAIN
(String) a part of the table value should match exactly the reference valueMATCHES
(String) the table value should match exactly the regular expression from the reference valueLOWER
(numbers) the table value should be strictly lower than the reference valueLOWER_EQUALS
(numbers) the table value should be lower than or equals to the reference valueGREATER
(numbers) the table value should be strictly greater than the reference valueGREATER_EQUALS
(numbers) the table value should be greater than or equals to the reference valueNULL
(String, boolean, numbers) the table value should be nullNOT_NULL
(String, boolean, numbers) the table value should not be null
ov filter remove
Removes the filter of the active table
List of commands - List of Manage tables commands
ov filter show
Show the filter window of the current table
List of commands - List of Access GUI dialogs commands
ov import node table
Transform node table into an Omics Visualizer table
-
importedColNames
(optional) StringComma-separated list of column names to import into the new Omics Visualizer table. Commas in the column names should be escaped.
-
importedNamespaces
(optional) StringComma-separated list of column namespaces to import into the new Omics Visualizer table. All the columns from those namespaces will be imported. Commas in the column namespaces should be escaped.
-
keyColName
(required) StringName of the shared column that identifies the row in the node table.
-
newTableName
(optional) StringThe name of the new Omics Visualizer table. A default generated name will be given if omitted.
-
srcName
(optional) String Default:source
, or the namespace if all the selected columns are from the same namespace.The name of the column containing the values source.
-
srcNetwork
(required) CyNetworkThe network containing the node table.
-
valuesName
(optional) String Default:value
The name of the column containing the values.
Nota Bene: importedColNames
and importedNamespaces
are optional, but at least one of them should be filled.
ov import node table srcNetwork="String Network" keyColName="display name" importedNamespaces="tissue" srcName="tissue" newTableName="String tissues"
This will import the columns from the "tissue" namespace of "String Network". The new Omics Visualizer table named "String tissues" will be connected with "String Network" with the column "display name". The column of "String tissues" that contains the name of the node table column is named "tissue".
List of commands - List of Manage tables commands
ov legend draw
Draw a legend
-
fitView
(optional) boolean Default:true
Should the view of the network be fitted to the content after the legend is added.
-
fontName
(optional) String Default:"SansSerif"
The font family.
-
fontSize
(optional) int Default:22
The font size.
-
includeInner
(optional) boolean Default:true
if there is an inner visualization,false
otherwiseShould the inner visualization included in the legend?
-
includeOuter
(optional) boolean Default:true
if there is an outer visualization,false
otherwiseShould the outer visualization included in the legend?
-
orientation
(optional) String Default:"HORIZONTAL"
The orientation of the legend.
Must be one of:"HORIZONTAL"
"VERTICAL"
-
position
(optional) String Default:"NORHT_LEFT"
The position of the legend.
Must be one of:"NORTH"
the legend will be on top of the network, centered"NORTH_LEFT"
the legend will be on top of the network, aligned left"NORTH_RIGHT"
the legend will be on top of the network, aligned right"SOUTH"
the legend will be at the bottom of the network, centered"SOUTH_LEFT"
the legend will be at the bottom of the network, aligned left"SOUTH_RIGHT"
the legend will be at the bottom of the network, aligned right"EAST"
the legend will be on the right of the network, aligned with the middle"EAST_TOP"
the legend will be on the right of the network, aligned top"EAST_BOTTOM"
the legend will be on the right of the network, aligned bottom"WEST"
the legend will be on the left of the network, aligned with the middle"WEST_TOP"
the legend will be on the left of the network, aligned top"WEST_BOTTOM"
the legend will be on the left of the network, aligned bottom
-
showTitle
(optional) boolean Default:false
Should the title be added to the legend?
-
title
(optional) String Default: the network nameThe title of the legend.
If you don't want any title, you give an empty title.
ov legend draw orientation="VERTICAL" position="EAST_TOP" title=""
This will generate a legend containing all the visualizations (inner and/or outer, if any) with no title. The legend will be drawned vertically at the right of the network, aligned with the top of it.
List of commands - List of Legends commands
ov legend hide
Hide a legend
List of commands - List of Legends commands
ov load
Load an Omics Visualizer table.
Returns the name of the newly imported OV table.
-
dataTypeList
(optional) StringList of column data types ordered by column index (e.g. "string,int,long,double,boolean,intlist" or just "s,i,l,d,b,il").
If not given, the types will be automatically detected. -
decimalSeparator
(optional) Character Default:'.'
Character that separates the integer-part (characteristic) and the fractional-part (mantissa) of a decimal number.
-
delimiters
(optional) String Default:","
if CSV,TAB
otherwiseSelect the delimiters to use to separate columns in the table.
Must be one of:","
" "
"\t"
";"
-
delimitersForDataList
(optional) Character Default:'|'
The delimiters between elements of list columns in the table.
Must be one of:'|'
'\'
'/'
','
-
file
(required) StringThe path to the file that contains the table or network to be imported.
-
firstRowAsColumnNames
(optional) boolean Default:true
Does the first imported row contains column names?
-
newTableName
(optional) StringThe title of the new table.
If no title is given, then a name will be generated, started by "Omics Visualizer Table" followed by a number. -
startLoadRow
(optional) int Default:1
The first row of the input table to load. This allows the skipping of headers that are not part of the import.
ov load file="/path/to/myfile.csv" newTableName="myData"
This loads the file "myfile.csv" into Omics Visualizer and create the corresponding table with the name "myData".
List of commands - List of Manage tables commands
ov palette list
List available palettes with their provider
ov retrieve
Retrieve a STRING network and connects it to the active OV table.
-
cutoff
(optional) Double Default: 0.4Confidence (score) cutoff. Must be between 0.0 and 1.0.
-
filteredOnly
(optional) boolean Default:true
Only filtered rows should be used to the query?
-
queryColumn
(required) StringColumn name of the Omics Visualizer Table.
-
selectedOnly
(optional) boolean Default:false
Only selected rows should be used to the query ?
-
species
(optional) StringName of the species to query.
-
taxonID
(optional) intIdentifier of the species to query.
Nota Bene: species
and taxonID
are optional, but at least one of them should be filled.
ov retrieve queryColumn="UniProt" taxonID=9606
This retrieves a STRING network for the active OV table. The query is made on UniProt values from the OV table for Human (taxon 9606
).
List of commands - List of Manage tables commands
ov retrieve show
Show the retrieve window of the current table
List of commands - List of Access GUI dialogs commands
ov table delete
Delete the active Omics Visualizer table
List of commands - List of Manage tables commands
ov table list
Get the list of Omics Visualizer tables
List of commands - List of Manage tables commands
ov table set current
Set the current Omics Visualizer table
-
tableName
(required) StringName of the OV table to set as active
ov table set current tableName="myData"
This sets the OV table named "myData" as the active one.
List of commands - List of Manage tables commands
ov version
Returns the current version of the app
ov viz apply inner continuous
Apply an inner visualization (pie charts) with a continuous mapping
-
attributes
(required) StringName of the column that contains the data you want to visualize.
-
chartSettings
(optional) StringComma separated list of enhancedGraphics settings.
Here is how the string should be formatted: "setting1:value1,setting2:value2" e.g. "arcstart:0,arcdirection:counterclockwise" -
colorMax
(optional) StringColor used in the gradient as the highest value.
-
colorMid
(optional) StringColor used in the gradient as the middle value.
-
colorMin
(optional) StringColor used in the gradient as the lowest value.
-
colorMissing
(optional) StringColor used for missing values.
-
filteredOnly
(optional) boolean Default:true
Use all the data (
false
) or only the filtered one (true
)? -
labels
(optional) StringColumn name of the table that should be used to label the data. By default no label is displayed.
-
paletteName
(optional) StringName of the palette to use as default colors. (See ov palette list)
-
paletteProviderName
(optional) StringName of the palette provider of the palette. (See ov palette list)
-
rangeMax
(optional) doubleMaximum value. Above this value, the same color will be applied.
-
rangeMid
(optional) doubleMiddle value. The colors will be a gradient from min to mid, and from mid to max.
-
rangeMin
(optional) doubleMinimum value. Below this value, the same color will be applied.
ov viz apply inner continuous attributes="log ratio" rangeMin=-5 rangeMax=5
This will create a pie chart visualization from the "log ratio" column, using the default palette colors with a range from -5 to 5.
ov viz apply inner continuous attributes="log ratio" rangeMin=-5 rangeMax=5 paletteName="Purple-Orange" paletteProviderName="ColorBrewer"
This will create a pie chart visualization from the "log ratio" column, using the specified palette colors with a range from -5 to 5.
ov viz apply inner continuous attributes="log ratio" rangeMin=-5 rangeMax=5 colorMin="blue" colorMid="white" colorMax="red"
This will create a pie chart visualization from the "log ratio" column, using the specified colors with a range from -5 to 5.
List of commands - List of Visualizations commands
ov viz apply inner discrete
Apply an inner visualization (pie charts) with a discrete mapping
-
attributes
(required) StringName of the column that contains the data you want to visualize.
-
chartSettings
(optional) StringComma separated list of enhancedGraphics settings.
Here is how the string should be formatted: "setting1:value1,setting2:value2" e.g. "arcstart:0,arcdirection:counterclockwise" -
colorMapping
(optional) StringComma separated values of mappings value:color. Special characters in values must be escaped.
-
colorMissing
(optional) StringColor used for missing values.
-
filteredOnly
(optional) boolean Default:true
Use all the data (
false
) or only the filtered one (true
)? -
labels
(optional) StringColumn name of the table that should be used to label the data. By default no label is displayed.
-
paletteName
(optional) StringName of the palette to use as default colors. (See ov palette list)
-
paletteProviderName
(optional) StringName of the palette provider of the palette. (See ov palette list)
ov viz apply inner discrete attributes="cluster" colorMapping="A:blue,B:orange,C:red"
This will create a pie chart visualization with the "cluster" column. The value "A" will be blue, "B" orange and "C" red. If there are more values in the "cluster" column, a default color from the default palette will be assigned.
List of commands - List of Visualizations commands
ov viz apply outer continuous
Apply an outer visualization (donuts charts) with a continuous mapping
-
attributes
(required) StringComma separated value list of attributes from the table you want to visualize the data. Special characters in the attributes must be escaped.
-
chartSettings
(optional) StringComma separated list of enhancedGraphics settings.
Here is how the string should be formatted: "setting1:value1,setting2:value2" e.g. "arcstart:0,arcdirection:counterclockwise" -
colorMax
(optional) StringColor used in the gradient as the highest value.
-
colorMid
(optional) StringColor used in the gradient as the middle value.
-
colorMin
(optional) StringColor used in the gradient as the lowest value.
-
colorMissing
(optional) StringColor used for missing values.
-
filteredOnly
(optional) boolean Default:true
Use all the data (
false
) or only the filtered one (true
)? -
labels
(optional) StringColumn name of the table that should be used to label the data. By default no label is displayed.
-
paletteName
(optional) StringName of the palette to use as default colors. (See ov palette list)
-
paletteProviderName
(optional) StringName of the palette provider of the palette. (See ov palette list)
-
rangeMax
(optional) doubleMaximum value. Above this value, the same color will be applied.
-
rangeMid
(optional) doubleMiddle value. The colors will be a gradient from min to mid, and from mid to max.
-
rangeMin
(optional) doubleMinimum value. Below this value, the same color will be applied.
-
transpose
(optional) boolean Default:false
Should the ring represents a column (
false
) or a row (true
)?
ov viz apply outer continuous attributes="log1,log2,log3" rangeMin=-2 rangeMax=2
This will create a donut chart visualization with three rings: "log1", "log2", and "log3". The colors are automatically assigned according to the default palette with a range from -2 to 2.
List of commands - List of Visualizations commands
ov viz apply outer discrete
Apply an outer visualization (donuts charts) with a discrete mapping
-
attributes
(required) StringComma separated value list of attributes from the table you want to visualize the data. Special characters in the attributes must be escaped.
-
chartSettings
(optional) StringComma separated list of enhancedGraphics settings.
Here is how the string should be formatted: "setting1:value1,setting2:value2" e.g. "arcstart:0,arcdirection:counterclockwise" -
colorMapping
(optional) StringComma separated values of mappings value:color. Special characters in values must be escaped.
-
colorMissing
(optional) StringColor used for missing values.
-
filteredOnly
(optional) boolean Default:true
Use all the data (
false
) or only the filtered one (true
)? -
labels
(optional) StringColumn name of the table that should be used to label the data. By default no label is displayed.
-
paletteName
(optional) StringName of the palette to use as default colors. (See ov palette list)
-
paletteProviderName
(optional) StringName of the palette provider of the palette. (See ov palette list)
-
transpose
(optional) boolean Default:false
Should the ring represents a column (
false
) or a row (true
)?
ov viz apply outer discrete attributes="cluster1,cluster2"
This will create a donut chart visualization with two rings: the "cluster1" and "cluster2" values. The colors for each value will be automatically generated using the default palette.
List of commands - List of Visualizations commands
ov viz remove inner
Remove the inner Visualization (pie charts) of the current network
List of commands - List of Visualizations commands
ov viz remove outer
Remove the outer Visualization (donuts charts) of the current network
List of commands - List of Visualizations commands
ov viz show inner
Show the inner visualization (pie charts) window of the current table
List of commands - List of Access GUI dialogs commands
ov viz show outer
Show the outer visualization (donuts charts) window of the current table