StringSearch identifies sequences by searching for character patterns such as "globin" or "human" in the sequence documentation.
In addition to the actual sequence data, GCG databases contain two additional types of data: sequence annotations, and definitions.
The annotations contain the complete documentation for each entry in the sequence database, including journal and author names, sequence features, comments, etc. The annotations appear at the top of sequences copied from a GCG database with the Fetch program.
The definitions contain a minimal amount of the annotations documentation for each entry: the name of the organism, the name of the gene, the sequence length, and usually the date. Definitions for the GenBank, EMBL, and SWISS-PROT databases also contain the primary accession number for the sequences.
The StringSearch program searches through either the definitions alone or the complete sequence annotations for text patterns that you specify. Annotations take much longer to search than definitions.
You can search for more than one text pattern in response to the program prompt with Human,Globin. StringSearch then finds all the entries that contain both human and globin. You can set StringSearch to show all the entries that contain either human or globin with Find entries: set to with ANY of the specified patterns.
Blank spaces are removed from the beginning of each pattern unless that pattern is enclosed in double quotes. For instance, specifying the pattern Globin shows all entries that contain globin, while specifying " Globin" excludes entries containing terms like myoglobin in which globin is not preceded by a space.
To specify a double quote (") as part of a pattern, use two double quote marks (""). To specify a comma as part of a pattern, enclose the whole pattern in quotes.
The output file from StringSearch is a list file. (See Chapter 2, Using Sequence Files and Databases of the User's Guide for more information.) Here is what the output from the example session looks like:
!!SEQUENCE_LIST 1.0 ! STRINGSEARCH from: GenEMBL:* October 22, 1998 11:38 ! searching for: "pseudo" .. Gb_ba:Ab000361 AB000361 Pseudomonas cichorii gene for D-Tagatose ... Gb_ba:Ab001577 AB001577 Pseudomonas sp. DNA for low specificity ... Gb_ba:Ab001722 AB001722 Pseudomonas stutzeri carbazole catabolic ... ////////////////////////////////////////////////////////////////////// Gb_sts:Ppu85462 U85462 Pseudomyrmex pallidus clone Psd2421AAG tri ... Gb_sts:Ppu85463 U85463 Pseudomyrmex pallidus clone Psd2427AAG tri ... Gb_sts:Ppu85464 U85464 Pseudomyrmex pallidus clone Psd2523CAC tri ... ! Sequences searched: 552323
The search is case insensitive.
You cannot assume that a text pattern search is exhaustive. The text you choose may not have been used by the data collectors. Worse yet, all databases contain errors -- the misspelling psuedo appears 14 times in the definitions for GenBank Release 108.0!
Hyphenation is particularly prone to inconsistent usage. A search for pseudogene would only be complete if pseudo-gene (or pseudo gene) were never used to refer to pseudogenes.
Using a nonspecific pattern such as pseudo to find sequences of pseudogenes will result in many false matches. (In the example session, there were 1570 instances of Pseudomonas, 321 instances of pseudoobscura, and 42 instances of pseudoautosomal out of 6237 total matches.) But restricting the search by setting Find entries: set to with ALL of the specified patterns with ALL of the specified patterns searching for pseudo,gene will miss pseudogene sequences whose definitions use terms like pseudoexon or the prefix pseudo- used with the name of the gene. It's usually better to use a less-specific search pattern and then edit the resulting list file to remove entries that you aren't interested in.
The conclusion is that a search with StringSearch can only tell you what is available and not what is not available.
The complete annotation search takes a lot of computing, but the search includes a lot of information, such as author and journal names, that is not found in the sequence definitions. You can speed up the search considerably by using a sequence specification like Primate:hum* to look only at the group of sequences in which you really expect the text pattern to be found.
You can set the parameters listed below from the command line. For more information, see "Using Program Parameters" in Chapter 3, Using Programs in the User's Guide.
Documentation Comments: doc-comments@gcg.com
Technical Support: help@gcg.com
Copyright (c) 1982, 1983, 1985, 1986, 1987, 1989, 1991, 1994, 1995, 1996, 1997, 1998, 1999 Genetics Computer Group Inc., a wholly owned subsidiary of Oxford Molecular Group, Inc. All rights reserved.
Licenses and Trademarks Wisconsin Package is a trademark of Genetics Computer Group, Inc. GCG and the GCG logo are registered trademarks of Genetics Computer Group, Inc.
All other product names mentioned in this documentation may be trademarks, and if so, are trademarks or registered trademarks of their respective holders and are used in this documentation for identification purposes only.