The GCG package of sequence analysis programs.

Current program status.

User interface.

Many of the GCG routines have an option to produce .png graphics files that you can download and display on your computer.  Before running the program producing the graphics file, run the command png.  This is an alternative to running the same program under seqlab, which would automatically export the file as an X-windows graphic.
 

Starting GCG.

• Place a file in your home directory named gcg.com with the following contents
source /share/apps/gcg/gcgstartup.ksh
gcg
set +o noglob
export LD_ASSUME_KERNEL=2.4.19

• Thereafter the command: source ~/gcg.com will start GCG.

Documentation.

Typical method of use.

• First get GCG running.  It will announce itself with a sign-on screen that lists numbers of databases.  GCG commands are given by typing at the ordinary linux command prompt.  Ordinary linux commands are still available
• Get the sequence of interest.  You will usually put it into your bioinf directory either by scp from your local machine, or by accessing a graphics desktop on bioinf by a graphics interface (such as VNC viewer) and using the Netscape browser to download the sequence directly from GenBank.  Get the sequence either in GenBank (aka. GenPept) or fasta format.  Do not download it as an html file.  Fasta format has a definition line starting with an >, followed by lines with nothing but sequence characters.
• Convert the sequence to GCG's internal format.  For fasta files, use the GCG command fromfasta.  For GenBank files use the GCG command fromgenbank.  Beware that the output file will be named after the sequence identifier found inside your file; not after the input file name.  Output files will be named with a .seq extension.  All other GCG commands will use the .seq files for input.  The above sequence of commands handles either protein or nucleic acid files.
• You can edit or complement a file using the GCG sequence editor named seqed.  Some commands have to be ended with a <control d>.  For example you type a <control d> to stop editing the header in seqed.  In seqed, you type a : to get a command line.  Issue the command help to see what other commands are available.
• To see what programs are available type genmanual from the linux command prompt, and explore.
• Most functions will lead you through interactively if you just type the function name.  There is usually more functionality available than meets the eye, so use genehelp if you want to find a way to modify the way the program works.
• If you want graphics output, you should first give the command png.
• Sequences can be put back to fasta format with the command tofasta.

Functions available in GCG.

1. General utilities to edit, complement, translate sequence, and to map restriction sites or peptidase cleavage sites.
• Similar functions exist in most all packages and at many remote sites on the web.
2. Facilities to retrieve sequences from GenBank, and to reformat them from a variety of formats.
• Comparable to using NCBI Entrez with at web browser, except sequences are automatically imported in GCG format
3. Automated and manual alignment programs.
• The automated alignment program, pileup, is less powerful that ClustalW, which can be run directly, or out of the SeaView editor.
• The SeqLab alignment editor is similar to SeaView, but SeaView has integrated phylogenetic analysis as well as integrated ClustalW.
• The manual alignment program lineup, is much like MSE, but MSE can handle more and longer sequences.
4. Blast, Fasta, and other search programs for databases.
• The NCBI programs at NCBI and installed locally in the Blast suite are more advanced.  The internal GCG databases fall out of date and should not be used.  The GCG local blast and psiblast programs can be redirected to search the daily updated local databases by the command line parameter -INfile2=$BLAST_DB/<database name>
5. Profile and motif construction and search facilities.
• HMMER itself is directly installed on bioinf.
• A more advanced package named SAM is also installed locally.
6. Implementations Neighbor Joining for constructing phylogenetic trees.
• Neighbor joining runs directly out of ClustalW.  PAUP, a more powerful package for phylogenetic analysis, is available on bioinf.  Phylip, which is extremely powerful but not very convenient is also installed.
7. A variety of algorithms for randomizing, scrambling, or randomly mutating sequence, for use in statistical testing.
8. Sequence assembly for sequencing projects.
• However, the capability to link to the original data is absent.
• The LaserGene Package under site-license is recommended for sequence assembly.  Vector NTI also has a package.  For very large projects, phred/phrap/consed is available.
9. A primer picking program.
• All of the commercial packages have comparable primer picking software.  We recommend the site-licensed Oligo6.0, because it has the most thorough documentation.
10. General utilities for counting codons and determining codon preferences.
11. RNA and DNA secondary structure prediction.
12. General facilities for creation of displays.
13. Prediction of protein features (helical wheel, transmembrane, secretion signals, coiled-coil regions, etc.).
14. Screening of internal repetitive sequence.

Troubleshooting GCG.

• Reinitialization.
• Command line GCG sometimes loses track of its initialization information and starts to give "not found" messages for its commands or for genhelp.
• Place a file in your home directory named gcg.com with the following contents

source /share/apps/gcg/gcgstartup.ksh
gcg
set +o noglob
export LD_ASSUME_KERNEL=2.4.19
 
• Thereafter the command: source ~/gcg.com will reinitialize GCG.
• ls fails to work with * as wild card.
• The gcg startup file executed at startup for most bioinf users inactivates wildcard usage.
• To reverse this, type set +o noglob.
• As indicated in the item above, you can put this fix in a gcg initialization file.
• Note: To prevent directory names from being expanded to directory contents use the -d option.
• For example, to find directory names starting with "p" , type ls -d p*, not ls p*
• The lookup command doesn't work right.
• Instead use NetFetch to get the sequence from NCBI Entrez, or use your browser to get it from NCBI Entrez.
• The commands refering to PAUP are not implemented.
• You may use the stand alone version of PAUP instead. See http://www.bioinformatics.uthscsa.edu/www/genomics.html#paup.
• Can't find a particular database.
• For blast and psiblast searching out of GCG, use -INfile2=$BLAST_DB/<database name> to refer to the daily updated NCBI libraries.  See notes about GCG use in local blast suite documentation.
• Other databases such as ProCite, Rebase, and Pfam are no longer being maintained by distribution from Accelrysis.  If there are up-to-date copies being maintained that GCG can access, then there will be a comment listed about them on the Bioinformatics GCG status page.  E-mail hardies@uthscsa.edu if you need some particular database to be updated.