Advanced Query Help Page

This page describes how to use the Pathway Tools Advanced Query Form. The form allows you to issue queries that retrieve information from EcoCyc and from pathway/genome datasets for other organisms. Each query retrieves a set of objects from a specified class according to constraints provided for attributes of that class.

WWW Browser Requirements: Because the Advanced Query Form uses modern Javascript features for updating the contents of the selector boxes, a modern WWW browser is needed. The form has been tested on Netscape 4.05 and Internet Explorer 4.0. However, it will not work on version 3.x of these browsers.

Selecting the Query Dataset and Query Class

The top part of the form is used for setting the context of the query. Each query applies to one dataset, and to one class of database objects (frames) associated with that dataset. These choices can be made with the topmost two selector boxes. The class defines a type of biological object. For example, you might choose to query the class of proteins associated with the E. coli dataset, or to query all genes associated with a different organism. The set of datasets available will vary at different sites.

Selecting the Query Connective

The third box is used for choosing the logical connective that is used to combine the results of each clause in the query to form the end result of the full query. If you choose AND, then the query will will return those results that are common to all of the query clauses; choosing OR will cause the query to return those results selected by any of the query clauses; choosing EXCLUSIVE-OR will return only those results selected by exactly one of the clauses.

Defining the Query Clauses

The actual query consists of a set of clauses. Each clause is a condition that is used to screen all objects in the class selected for the query. The condition refers to slots (attributes) of the objects.

For example, if the class of small molecules (called Compounds) is selected, each clause is evaluated against every small molecule in the database for the selected dataset. Every class has a different set of slots; each clause refers to one slot, and is selected using a box within the clause. Compounds, for example, contain slots such as Molecular-Weight and Names whose values are a numerical molecular weight and one or more names for the chemical compound, respectively. A clause can be used to compare the values of a slot for each compound against a value supplied by the user.

The main part of the form contains 5 identical sections for formulating clauses of the query. At least one clause should be filled in for each query -- the remaining sections should be left blank.

The selector boxes within a clause are used to build up the statement of the condition to form a readable phrase such as, "The value of slot molecular-weight is < 54". The meaning of those boxes is as follows.

  1. The first selector box allows querying directly for the value in a given slot, or alternatively, for the number of values that a slot contains, because some slots can contain more than one value. This latter setting can be used to ask whether a given slot has any values at all.

  2. The second selector box should be used to specify the name of the slot one would like to query.

  3. The third selector box allows negation of the slot query.

  4. The fourth selector box should be used to specify the comparison operator that will be used for matching values in the slot to the query value that the user enters into the form. The allowed operators are determined by the type of values that are allowed to reside in the slot that was selected under (2). The main distinction is between numbers that allow operations such as greater-than and less-than, and text strings that allow substring searches.

  5. Last is a text box in which the user enters a query value for comparison purposes, which together with the operator will determine whether a frame satisfies the slot query. The values that can be entered are numbers and text strings. For slots that take boolean values, you may use the strings TRUE and FALSE for comparison purposes. The values of some slots are frame IDs, so you may enter frame IDs as strings in this box.

If in any of the slot queries the choice of slot name or operator is not made, or no query value is entered in the text box, that particular slot query is ignored from the final results.

Defining the Query Results

The lower part of the form contains a multiple selection box that allows you to choose what slot values should be included in the table of frames that will be returned as the query result. If no slots are chosen here, the table will only contain the name of each result frame.

Additionally, a check box allows inclusion of the frame ID (the internal database unique identifier) in the results table. The frame ID is used in URL links to EcoCyc objects.

Finally, at the bottom of the form, clicking on the Submit button will send the query to the server, and will return the results in a table. Each result row corresponds to one frame that satisfied the query. The columns of the table list the values of the slots that were chosen by the user. The first column shows the "common name" of the frame if one is available; the second column optionally shows the frame ID, if this choice was made above.

Additionally, the result page includes the original query.

Example Queries

1. As a simple query to start with, let's assume we would like to ask for all compounds in E. coli that contain "tryp" in one of their names or synonyms. The slot we would query is called NAMES. Because we are querying for only one slot, the connective chosen does not matter. The query that will answer our question is: 

In dataset [E. coli], in class [Compounds] :
AND
     The [value] of slot [NAMES] [is] [superstring of] ["tryp"]

In the early 1999 version of EcoCyc, this query will return 11 compound frames.

We can further narrow down the results by additionally asking for only those compounds that are larger than 200Da. To achieve this, we can query an additional slot that stores the MOLECULAR-WEIGHT. We will need to connect the two slot queries with AND, like this: 

In dataset [E. coli], in class [Compounds] :
AND
     The [value] of slot [NAMES] [is] [superstring of] ["tryp"]
     The [value] of slot [MOLECULAR-WEIGHT] [is] [larger than] ["200"]

This query reduces the number of returned frames to 8.

2. As a more complicated example, let us assume we would like to investigate the usage of the term "synthase" and "synthetase". Do they tend to be used interchangeably, or is there a difference in meaning and usage ? The class of frames we need to query is Enzymatic-Reactions, because this class encodes enzyme names.

First, we can just query for all the frames that contain either or both of these terms, and so our connective would be OR: 

In dataset [E. coli], in class [Enzymatic Reactions] :
OR
     The [value] of slot [NAMES] [is] [superstring of] ["synthase"]
     The [value] of slot [NAMES] [is] [superstring of] ["synthetase"]

This query returns 177 frames. This query, however, will not tell us whether some of these frames use both terms at the same time. If we change the connective to AND, we will be able to see that 42 frames are returned.

To find the frames which use these terms unambiguously, i.e. uniquely, we can change the connective to EXCLUSIVE-OR. This will return 135 frames. Looking at the frames returned in these two sets could tell us something about whether the uses of these terms are very strict or rather loose.

To find all frames that only use one term, but not the other, we can modify the query as follows, using negation of one term: 

In dataset [E. coli], in class [Enzymatic Reactions] :
AND
     The [value] of slot [NAMES] [is] [superstring of] ["synthase"]
     The [value] of slot [NAMES] [is not] [superstring of] ["synthetase"]

We find that 71 frames only use "synthase", and 64 use "synthetase". These two numbers add up the what the EXCLUSIVE-OR query returned, as we expected. It appears that the usage of these two terms is approximately balanced, and used relatively interchangeably.

Return to SGD
Send a Message to the SGD Curators