Bug Fixes
You will find the following information in this section:
- Program Bug Fixes
- Package-Wide Bug Fixes
Program Bug Fixes
-
-
Editing
-
-
-
SeqEd
-
Problem: You were unable to enter the tilde (~) to indicate NOT syntax when you specified a pattern to find in the sequence you were editing.
Update: You can enter the tilde (~) to indicate NOT syntax when you specify a pattern to find in the sequence you are editing.
-
-
Mapping
-
-
-
Map
-
Problem: The reading frames for the forward strand protein translations were not changed if you specified a beginning position other than the first position in the sequence. For instance, frame a always began at the first position in the sequence rather than at the beginning of the selected range. Similarity, reading frames for the reverse strand protein translations were not changed if you specified an ending position other than the last position in the sequence. For instance, frame f always began at the last position in the sequence rather than at the ending of the selected range.
Update: If you specify a beginning position other than the first position in the sequence, the forward strand reading frames are changed so that the a frame starts at the beginning of the selected range. If you specify an ending position other than the last position in the sequence, the reverse strand reading frames are changed so that the f frame starts at the end of the selected range.
-
MapSort
-
Problem: Unless you added the undocumented -PROtein command-line parameter, the program recognized all input sequences as nucleic acids.
Update: The program determines the sequence type from the Type: field of the input sequence (see Appendix VI in the Program Manual) in the same way the sequence type is determined by other Wisconsin Package analysis programs.
Problem: If you looked for silent restriction sites by adding -SILent to the command line, and you specified a sequence range that did not extend to the end of the sequence, the program sometimes crashed if a silent restriction site was identified at the very end of the sequence range.
Update: The program identifies silent restriction sites at all appropriate sites in the sequence without crashing.
-
Prime
-
Problem: Prime rarely calculated incorrect annealing scores between two primers. In addition, Prime also rarely calculated incorrect annealing scores between a primer and possible false priming sites on the template sequence when you specified either -ALLANNEALTemplate or -ENDANNEALTemplate on the command-line. Because of this, some primers were incorrectly rejected from consideration and others might not have been ranked at their appropriate positions in the output list.
Update: Prime correctly calculates annealing scores between two primers or between a primer and possible false priming sites on the template sequence.
Problem: If you specified an input file of primer sequences with the -PRImers command-line parameter, and you also specified either -ALLANNEALTemplate or -ENDANNEALTemplate on the command line, Prime sometimes calculated incorrect annealing scores between the primers and possible false priming sites on the template sequence. Because of this, some primers were incorrectly rejected from consideration and others may not have been ranked at their appropriate positions in the output list.
Update: Prime correctly calculates annealing scores when you specify an input file of primer sequences and you also specify either -ALLANNEALTemplate or -ENDANNEALTemplate on the command-line.
Problem: If you chose an alternative output filename and you specified -BATch on the command line, the output filename you chose was ignored and the output file was given the default name.
Update: Prime writes its output to the file you name when you specify -BATch on the command line.
-
-
Comparison
-
-
-
BestFit and Gap
-
Problem: If you aligned sequences contained in an MSF file, the output listed the name of the MSF file and not the names of the sequences.
Update: If you align sequences in an MSF file, the output lists the names of the sequence.
-
FrameAlign
-
Problem: If either input sequence was contained in an MSF file, the output alignment listed the name of the MSF file instead of the name of the sequence within the MSF file.
Update: The output alignment lists the name of the sequence within the MSF file.
-
Overlap
-
Problem: If you specified a group of sequences within an MSF file as input, each overlap listed the name of the MSF file instead of the names of the overlapping sequences.
Update: Each overlap lists the names of the overlapping sequences within the MSF file.
-
-
Database Searching
-
-
-
BLAST
-
Problem: If you specified a database to search on the command line with an expression like -INfile2=MyDir:nuc and a database of the same name, but in a different directory, was also found in one of the database menu files (blast.rdbs, blast.ldbs, or blast.sdbs), the database in the menu file was searched.
Update: The database you specify on the command line is searched.
Known Problem: With the increasing sizes of the databases, it is possible that you will not have enough memory to run a local BLAST search. We have already seen this happen on some machines when a nucleotide sequence is used to search GenEMBL. The program terminates before the search begins and the program output file contains an "out of memory" error message.
Possible Work-arounds: Increase the limits on your account by typing % unlimit before running BLAST. This may help if the query sequence is short. If this doesn't work, search each strand of a nucleotide query sequence separately using the commands % blast -TOPstrand and % blast -BOTtomstrand. If this doesn't work, speak to your system manager about either partitioning the large BLAST database into two or more smaller BLAST databases or increasing swap space.
-
LookUp
-
Problem: If you specified a query on the command line without a parameter (e.g. % lookup Smithies), the program stopped after complaining that no valid libraries exist.
Update: If you specify a query on the command line without a parameter, the program behaves as if you had used the -ALLtext parameter with the query and proceeds normally.
Problem: If your output list file contained comments exactly eighty characters in length, you could have problems if you tried to use the file as input to other Wisconsin Package programs. If you tried to do so, all sequence entries following those with long comment lines were skipped. Also, the long comment lines may have been missing a character at the end of the line.
Update: If your output list file contains comments exactly eighty characters in length, you can use the file as input to other Wisconsin Package programs without problem. Also, the long comment lines are no longer missing any part of the comment.
Problem: If you searched for sequences in SWISS-PROT, the program occasionally crashed.
Update: The program finds the appropriate matches in SWISS-PROT to your query and completes normally.
Problem: If you searched for some authors associated with sequence entries in SWISS-PROT, no matching entries were found.
Update: You can search for any authors associated with sequence entries in SWISS-PROT and the appropriate matching entries are reported.
Problem: If you searched for feature names containing apostrophes (e.g. 5'UTR) in GenBank, the program displayed a message claiming a syntax error in the feature name, and matching entries were not found.
Update: You can search for feature names containing apostrophes in GenBank, and the program reports the appropriate matching entries.
Problem: If you searched PIR and selected fragment output by adding -FRAgments to the command line, the program crashed.
Update: The program stops normally after displaying a message reporting that fragment output from PIR is currently unavailable.
-
FastA and TFastA
-
Problem: If you added -PAMfactor to the command line when searching with a nucleic acid query sequence, the command-line parameter was ignored. Instead of using a scoring matrix for the calculation of initial diagonal scores as you requested, the program used a constant factor for each match.
Update: When you add -PAMfactor to the command line in a nucleic acid search, the program uses a scoring matrix for the calculation of initial diagonal scores.
Problem: If either the query or matching search set sequence contained more than six letters, only the first six were displayed to the left of the sequences in the alignment output.
Update: There is now space for up to twelve letters in the sequence names to the left of the sequences in the alignment output.
-
FrameSearch
-
Problem: If you specified a search set that contained no valid sequences, and you added -Default to the command line, the program automatically substituted SwissProt:* (for a nucleotide query) or EST:* (for a protein query) as the search set.
Update: If you specify a search set containing no valid sequences and you add -Default to the command line, the program displays an error message and stops.
Problem: The program occasionally miscalculated the "Percent similarity" reported for each alignment when you specified an identity threshold with the -PAIr command-line parameter.
Update: The program correctly calculates the "Percent similarity" reported for each alignment under all circumstances.
Problem: If you ran FrameSearch with -BATch on the command line, and you specified multiple query sequences as input, the program used the sequence length of the first query sequence as the ending position for all subsequent query sequences.
Update: If you specify multiple query sequences as input, the program uses the entire length of each sequence in the search (unless you specify -BEGin or -END on the command line).
-
FindPatterns
-
Problem: If you allowed mismatches between the pattern and sequence by specifying -MISmatch on the command line, and a mismatch occurred in that portion of the pattern containing OR matching, the program sometimes missed finding all the appropriate matches.
Update: All appropriate matches are found when you allow mismatches and your pattern contains OR matching.
Problem: If you specified -PERFect on the command line in a search of nucleotide sequences, only the forward (top) strand of each nucleotide sequence was searched for matches to each pattern.
Update: Both strands of each nucleotide sequence are searched for perfect (non-ambiguous) matches to each pattern.
Problem: If no matches were found, then no output file was created.
Update: If no matches are found, then an output file is written indicating this result.
Problem: If you specified patterns containing spaces on the command line using the -PATterns command-line parameter, no matches were reported. However, if you specified patterns containing spaces in response to the program prompt, the spaces were automatically removed before searching and the appropriate matches were reported.
Update: If you specify patterns containing spaces on the command line, the spaces are removed before searching and the appropriate matches are reported.
Problem: If you searched for patterns in sequences contained in an MSF file, the output listed the name of the MSF file and not the names of the sequences in which the patterns were found.
Update: The output lists the names of the sequences within the MSF file in which the patterns are found.
-
StringSearch
-
Problem: If you entered a character pattern consisting of two words separated by more than one space, and you added -BATch to the command line, the program removed all but one of the spaces separating the words before searching for the pattern.
Update: StringSearch no longer removes any spaces between words in a character pattern when you add -BATch to the command line.
Problem: If you created a list file of sequence names with LookUp, and then used that list as input to StringSearch for a definitions search, StringSearch was unable to find matches to any text patterns.
Update: You can use a LookUp output file as input to StringSearch for a definitions search, and appropriate matches to the text patterns you specify are found.
-
GCGToBLAST (formerly ToBLAST)
-
Problem: If you tried to create a BLAST-searchable database from nucleic acid sequences containing X sequence symbols, GCGToBLAST complained that X was an invalid nucleic acid code because BLAST does not recognize the X as a nucleic acid ambiguity code.
Update: GCGToBLAST converts each occurrence of X (or x) in nucleotide sequences into an N (or n) nucleic acid ambiguity symbol in the BLAST-searchable database.
-
-
Multiple Sequence Analysis
-
-
-
PileUp
-
Problem: You could not increase the combined length of all gaps that could be added to each sequence in the alignment with the -MAXGap command-line parameter unless you also decreased the maximum segment length of each input sequence with the -MAXSeg command-line parameter.
Update: When you increase the combined length of all gaps that can be added to each sequence in the alignment with the -MAXGap command-line parameter, the maximum segment length of each input sequence is automatically reduced so that the sum of the maximum segment length and the maximum gap length is equal to 7,000.
-
LineUp
-
Problem: On the command line, you were unable to use a normal MSF sequence specification like ^&^ myseqs.msf{*}. Instead, you had to use a LineUp-specific syntax like % lineup myseqs.msf.
Update: LineUp accepts any single or multiple sequence specification on the command line using the normal syntax.
Problem: If your your local directory contained a set.keys file specifying keyboard key redefinitions for editing nucleotide sequences and the first sequence you entered into LineUp was a nucleotide sequence, the key redefinitions were ignored.
Update: If the first sequence entered into LineUp is a nucleotide sequence, the keyboard keys are redefined according to the specifications in the Set.Keys file in your local directory.
Problem: When you used the ZIp command to align a protein sequence to an existing protein consensus sequence, the program could propose meaningless alignments involving the reverse complement (-) strand of the protein sequence.
Update: The program does not propose alignments that involve the reverse complement strand of a protein sequence.
Problem: You were unable to enter the tilde (~) to indicate NOT syntax when you specified a pattern to find in the sequence you were editing.
Update: You can enter the tilde (~) to indicate NOT syntax.
-
PlotSimilarity
-
Problem: If you added -PROFile to the command line and specified non-profile input, the program displayed an error message and continued prompting for additional parameters. If you responded to the additional program prompts, the program crashed.
Update: If you add -PROFile to the command line and specify non-profile input, the program displays an error message and stops.
Problem: If you tried to plot the running average similarity among the sequences in an alignment longer than 11,500 symbols, the program rejected any density for the plot you specified in response to the program prompt and continued to prompt you for a new density. You had to use <Ctrl>C to exit the program.
Update: You can plot the running average similarity among the sequences in an alignment of any length. However, since the entire plot must fit on a single page, the plot becomes more difficult to read as the length of the alignment increases.
-
ProfileMake
-
Problem: If you entered more than the maximum limit of 100 sequences as input, the program displayed an error message for each additional sequence it tried to read.
Update: You can now enter up to 5,000 sequences as input. If this limit is exceeded, the program stops immediately.
-
ProfileSearch
-
Problem: If you specified a search set that contained no valid sequences, and you added -Default to the command line, the program automatically substituted SwissProt:* (for a protein profile) or EMBL:* (for a nucleotide profile) as the search set.
Update: If you specify a search set containing no valid sequences and you add -Default to the command line, the program displays an error message and stops.
Problem: If you specified a search set containing sequences whose lengths were all very similar, the program crashed while trying to normalize the scores.
Update: The program no longer tries to normalize the scores when the lengths of all the sequences are very similar.
-
-
Evolutionary Analysis
-
-
-
GrowTree
-
Problem: If you chose to reconstruct a tree from a distance matrix using the UPGMA method, the branch lengths in the output tree were incorrect.
Update: The branch lengths in the output tree are correct.
-
-
Pattern Recognition
-
-
-
CodonPreference
-
Problem: If you specified a range of the input sequence to analyze, and you chose to reverse that specified segment either by adding -REVerse to the command line or by responding to the program prompt, the entire sequence was first reversed and then the specified segment was selected from the reverse sequence strand.
Update: If you specify a range of the input sequence to analyze, and you choose to reverse the specified segment, the specified segment is first chosen from the forward sequence strand and then this segment is reversed. This is consistent with the behavior of other Wisconsin Package programs that offer you the option to analyze the reverse strand of the input sequence.
-
Repeat
-
Problem: If a repeat was longer than 55 bases or residues, a length of 55 was reported in the output file to the right of the repeat alignment.
Update: If a repeat is longer than 55 bases or residues, the correct length is reported in the output file to the right of the repeat alignment. However, only the first 55 bases of the repeat are actually displayed in the alignment.
-
CodonFrequency
-
Problem: If you tabulated the codon usage of a single sequence specified in a list file, any attributes associated with that sequence were ignored unless you added -Default to the command line.
Update: If you tabulate the codon usage of a single sequence specified in a list file, any begin, end, and strand attributes associated with that sequence are used as default input values by the program.
-
-
Protein Analysis
-
-
-
PlotStructure
-
Problem: If you create a figure file of the 1-dimensional panel graph plot by specifying -FIGure on the command line, and you also specified a font for all text characters in the plot using -FONT, the program crashed.
Update: You can specify both -FIGure and -FONT on the PlotStructure command line and create a figure file of a 1-dimensional panel graph plot without problem.
-
-
Translation
-
-
-
Translate
-
Problem: If you ran Translate noninteractively by specifying multiple sequences as input on the command line or by adding -Default to the command line, the program occasionally gave the output file a nucleotide sequence type. This occurred when the translated sequence contained only amino acid symbols that could be recognized as IUPAC-IUB nucleotide ambiguity symbols.
Update: Translate always writes an output file with a protein sequence type.
Problem: If you translated a single sequence specified in a list file, any attributes associated with that sequence were ignored unless you added -Default to the command line.
Update: If you translate a single sequence specified in a list file, any begin, end, strand, and join attributes associated with that sequence are used as default input values by the program.
-
BackTranslate
-
Problem: If your protein input sequence contained gap characters and you selected one of the table of back-translations menu choices (option a or b), then the table of back translations contained three periods in a row (...). Even though the output file also contained a GCG sequence appended after the table, the file was not recognized as a GCG sequence file by analysis programs.
Update: If your protein input sequence contains gap characters and you select one of the table of back-translations menu choices, the gap characters are back-translated to three tildes in a row (~~~). GCG analysis programs recognize the output file, also containing a GCG sequence appended after the table, as a GCG sequence file.
-
-
Manipulation
-
-
-
Assemble
-
Problem: If you interactively assembled a user sequence and then chose to G)et segments from another sequence, but specified the same user sequence again, the program didn't recognize the sequence the second time.
Update: You can repetitively assemble fragments from a single sequence by repetitively specifying the same sequence name in response to the program prompt.
Problem: If you assembled a single sequence specified in a list file, any attributes associated with that sequence were ignored unless you added -Default to the command line.
Update: If you assemble a single sequence specified in a list file, any begin, end, strand, and join attributes associated with that sequence are used as default input values by the program.
-
Simplify
-
Problem: If you entered a multiple sequence specification as input (for example, a sequence specification with an asterisk (*) wildcard) and that multiple sequence specification referenced only a single sequence, the program didn't read the sequence.
Update: You can enter any valid single or multiple sequence specification as input, even if the multiple sequence specification references a single sequence. This is consistent with the behavior of other Wisconsin Package programs that accept either a single or multiple sequences as input.
-
-
Display
-
-
-
Publish
-
Problem: If you selected one of the translation menu choices with numbering (option F or G), the translations may have been numbered incorrectly in several instances. For example, if the translation began in the middle of a row of the nucleotide sequence, it was numbered as if it began at the beginning of the row. If you chose three-letter translations (option F), and an amino acid began at the end of one row and stopped at the beginning of the next row, the numbering was incorrect. If you selected more than one discontinuous translation range (e.g. translations of exons separated by introns), they were numbered as if the entire sequence from the beginning of the first range to the end of the last range had been translated.
Update: The translation numbering is now correct in Publish.
-
-
Sequence Exchange
-
-
-
Reformat
-
Problem: If your attempt to reformat a sequence did not succeed, the input file was deleted.
Update: If your attempt to reformat a sequence does not succeed, the input file is not deleted.
Problem: If you reformatted a scoring matrix found in a directory other than your local directory by specifying a directory path along with the filename, the output scoring matrix was written to that same directory by default.
Update: If you reformat a scoring matrix found in a directory other than your local directory, the output scoring matrix is written to your local directory by default.
-
FromStaden
-
Problem: Staden-format nucleotide sequences containing lowercase IUPAC-IUB sequence characters were converted to periods (.) in the GCG-format output sequence files.
Update: Staden-format nucleotide sequences containing lowercase IUPAC-IUB sequence characters are unchanged in the GCG-format output sequence files. Appendix III of the Program Manual contains an updated list of the mappings between Staden and GCG sequence characters.
-
FromFasta
-
Problem: In a FastA-format input file, if the documentation following the sequence name contained two adjacent periods (..), the output sequence file had two lines containing two adjacent periods. This file was not recognized as a GCG sequence file by analysis programs.
Update: If the documentation following the sequence name contains two adjacent periods (..) in a FastA-format input file, the program inserts a blank space between the periods (. .) in the output file. GCG analysis programs will then recognize the output sequence file as a GCG sequence file.
Problem: If the documentation following the sequence name for a FastA-format sequence was longer than 511 characters, the program crashed.
Update: If the documentation following the sequence name for a FastA-format sequence is longer than 511 characters, it is written as several shorter documentation lines in the GCG-format output sequence file and the program completes normally.
-
Package-Wide Bug Fixes
-
-
Graphics
-
Problem: In any Wisconsin Package plotting program, if you used the -COPies command-line parameter to specify more than one copy for a plot sent to a PostScript device, only a single copy was plotted.
Update: All of the plot copies you specify with the -COPies command-line parameter are actual plotted.
-
List Files
-
Problem: In any program that recognizes the begin: and end: sequence attributes in a list file, if you specified -BEGin or -END on the command line without any value, the program ignored that command-line parameter.
Update: In any program that recognizes the begin: sequence attribute in a list file, if you specify -BEGin without any value, the program uses a beginning position of 1 for each sequence; beginning positions specified for individual sequences in the list file are ignored. In any program that recognizes the end: sequence attribute in a list file, if you specify -END without any value, the program uses the end of each sequence as the ending position; ending positions specified for individual sequences in the list file are ignored.
[ Program Manual | User's Guide | Data Files | Databases ]
Documentation Comments: doc-comments@gcg.com
Technical Support: help@gcg.com
Copyright (c) 1982, 1983, 1985, 1986, 1987, 1989, 1991, 1994, 1995, 1996, 1997 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.
