How To and FAQs
How To ...
- Submit Batch Jobs
- Make High Resolution Stereo Images Using InsightII (About InsightII)
- Print Molecule in color
- Run applications and get documentation
- How do I change my default shell from C shell to Korn shell on the Alpha?
- How do I interpret DSSP structural assignments?
- How do I use fasta to search my own database?
- How do I run Fastlink?
- How can I run multiple Blast searches?
- How do I run SeqWeb programs?
- How do I automatically initialize the GCG programs?
Frequently Asked Questions (FAQs)
Click on the links below to view the actual files.
Let us demonstrate for the molecule new0 , a plain benzene molecule:
First, generate the files:
- new0.car ==> Coordinate file
- new0.mdf ==> Molecular Topology file
- new0.inp ==> Input file to control the run
and any other relevent file using InsightII
Now, execute the following command (% denotes that this is a command that you need to execute). Characters you type are RED.
% source /usr/local/biosym/cshrc_980
This command sets up the appropriate ENVIRONMENT variables. Now execute the commands as shown below :
% discover new0
Which potential function do you wish to use?
1) library /usr/local/fbscapp/biosym/980/irix6m4/biosym_lib/cvff.bin
2) library /usr/local/fbscapp/biosym/980/irix6m4/biosym_lib/amber.bin
3) library /usr/local/fbscapp/biosym/980/irix6m4/biosym_lib/cff91.bin
4) library /usr/local/fbscapp/biosym/980/irix6m4/biosym_lib/pcff.bin
5) other
>> ENTER 1, 2, 3, 4 or 5, ( = 1) 1
>> ENTER the nice number for running discover as a background process.
The default is 19
>>
>> Enter number of processors - default is 8
>> 2
>> Do you wish to start discover now?
The default is YES
>> n
new0.csh created but not run
Now submit the job using the command
% qsub new0.csh
In InsightII,
- Load the molecule.
- Make the necessry modifications like, changing the background color, Annotations etc.
- In the "File" pulldown menu,
- Choose "Export_Image" and this pops up the EXPORT_IMAGE winow. In this window, choose the options:
- Image format: SGI_RGB
- Color Space: RGB
- Image Size: USER_DEfined
- Specify Dest Size: By_Scale
- Scale :<1.0>
- Image Quality: 600_dpi or 300_dpi
- Output File Name: whatever.rgb
- Move the curson into the Window containing the molecule, it looks like a pen now.
- Click and hold the left mouse button to draw and select the area from the screen which you want to print out.
- Click execute in the EXPORT_Image window.
- Use the application software Tekgui to printout the image.
To print images in Stereo
IF YOU HAVE MORE THAN ONE OBJECT IN YOUR SYSTEM, REMEMBER TO ASSOCIATE THEM INTO A SINGLE ASSEMBLY BEFORE GOING ANY FURTHER.Characters you type are RED.
- Follow step above to printout an image named "left.rgb"
- In InsightII, use the "Rotate" command under the "Transformation" menu to rotate the molecule by 3 degrees about the Y-Axis.
- Now follow steps above to printout an image named "right.rgb"
- These two images can be now combined into a single stereo image:
Go into the unix prompt in SGI, and execute the following command,
assemble 2 1 stereo.rgb left.rgb right.rgb - Use Tekgui to printout the stereo image.
Note: You may have to adjust the scaling of the image in Tekgui to get proper stereo.
Using Tekgui to print images
- If /usr/local/bin/ is not in your path, then execute the following command to include it:
set path=($path /usr/local/bin) - Now execute the command, tekgui
- This would bring up the Tektronix Graphical User Interface
- In the File menu, click on Open and open your image file.
- In the File menu, click on Printers and choose the appropriate printer.
- In the File menu, click on Print. Then, click on Setup or Layout and make appropriate choices.
- Click Print and your printout is on its way.
To select the Korn shell on the Alpha nciaxp:
Characters you type are RED.
You must type these characters exactly as shown.
- log in to nciaxp Alpha
- % cd
- % chsh
Old shell: /bin/csh
New shell: /bin/ksh - % cp /seq/app/documents/.profile .
- % cp /seq/app/documents/.kshrc .
- log out of nciaxp Alpha
Characters you type are RED.
Running Application Software
Application on SGI- In the table containing the list of softwares, click on the software name . Follow the instructions on the the getting started section on the page that appears.
- After logging in, execute the command analysis.
- Type bioinfo to get information about changes/enhancements and new programs available for sequence analysis.
Documentation on Application Software
In the table containing the list of softwares,- the "Man Page available" column and "Web Page Available" column contains appropriate browser links, click on them.
- Also, click on the software name. On the page that appears, the getting started section contains the appropriate pointers to local documentation and Web Site are available.
In case of difficulty, please contact Scientific Application Support.
H - 4-helix (alpha-helix)
B - residue in isolated beta-bridge
E - extended strand, participates in beta-ladder
G - 3-helix (3-10 helix)
I - 5-helix (pi helix)
T - H-bonded turn
S - bend
(In case of structural overlaps, priority is given to structure first in this list)
For details see: TABLEII in : Kabsch and Sander, Biopolymers, Vol. 22, pp. 2577-2637 (1983)
Characters you type are RED.
Example of using fasta with your own nucleotide database:- If you have a data file : /users/group/username/nuc.seq
- Create a file like: /users/group/username/data.list with a line like: NUC data$1A/users/group/username/data.list 1
The "1" at the end of the above line assumes your database is in Genbank format. If not then substitue the proper number from the list:
0 Pearson/FASTA (>SEQID - comment/sequence)
1 Uncompressed Genbank (LOCUS/DEFINITION/ORIGIN)
2 NBRF CODATA (ENTRY/SEQUENCE)
3 EMBL/SWISS-PROT (ID/DE/SQ)
4 Intelligenetics (;comment/SEQID/sequence)
5 NBRF/PIR VMS (>P1;SEQID/comment/sequence)
6 GCG (version 8.0) Unix Protein and DNA (compressed)
11 NCBI Blast1.3.2 format (unix only)(If you have a peptide database then substitute a "$0" for the "$1" above.)
- Run fasta with a command like (asking for the top 20 hits and alignments):
% /seq/app/fasta/fasta -Q -b 20 -d 20 -l /users/group/username/nuc.seq \ query.seq A > /tmp/username/query.fastaThis will search with the sequence in query.seq and put the output in /tmp/username/query.fasta
Characters you type are RED.
To run Fastlink Programs : ilink lodscore linkmap mlink
You need a pedigree ".pre" file and a parameter file. Type:
$ analysis
$ linkage
$ makeped
Makeped will convert your pedigree ".pre" file to a ".dat" file. Type:
$ lcp
Answering appropiately, selecting the program you want to run. When finished Type:
$ qpedin
This will submit an nqs batch job selecting the FASTLINK version of: ilink lodscore linkmap mlink
Characters you type are RED.
Doing multiple BLAST searches on our Alpha:
This method builds a script that is submitted to an nqs queue. It runs GCG's NetBlast repeatedly. It was created for screening of many sequences against EST and doesn't have a polished user interface yet.
Works best if you have all your query sequences (in GCG format) in a directory with names that have a .seq extension.
Type :
% analysis
% gcg
% ncbi
% mblast
This makes a script that you then submit to an NQS queue with:
% qsub multiblast
This default mode will search with each *.seq file in your current directory vs. EST, reporting 15 hits with alignments for each query. Each query must not contain more than 25% N and/or X, using blastn.
You can modify the request with 7 command line parameters:
mblast [1] [2] [3] [4] [5] [6] [7]
Where:
1 - directory path to query sequences
2 - extension of query files [ie. seq]
3 - database to search
4 - number of hits to report and align for each query
5 - percent (in decimal) max allowed N and/or X bases
6 - minimum length of query
7 - blast program to use
For example if I'm in my tmp directory (/tmp/gws) and have 3 queries (one.seq, two.seq, three.seq). Typing mblast is equivalent to typing:
mblast /tmp/gws seq est 15 0.25 0 blastn
When it is finsihed I will get these files in /tmp/gws:
one.n_est
two.n_est
three.n_est
Each being an individual blast search result file. If I want to see 40 hits I would Type:
% mblast . . . 40
Note: the 3 periods as place holders for the 1st 3 default parameters.
If my query files had extension .nuc (one.nuc, two.nuc, three.nuc) and I wanted to see 50 hits vs EST's I would Type:
% mblast . nuc . 50
If my query files had extension .nuc (one.nuc, two.nuc, three.nuc) and I wanted to see 50 hits vs NR I would Type:
% mblast . nuc nr 50
The WEB interface to the GCG programs is now available! From your browser (Netscape or Explorer) go to: http://seqweb.ncifcrf.gov/.
You must login in with a username and password. Your SeqWeb username is the same as your Alpha (nciaxp) username. ----> Your seqweb password is NEW (*****NOT your Alpha password). We prefer that you CALL for your SeqWeb password:
Call for your SeqWeb password, or if you have questions, contact: Gary W. Smythers 301-846-5778 or Bob Stephens 301-846-5787 or The Help Desk 301-846-5555.
Call to get your password, go to the SeqWeb page, login, and select the bottom line "Preferences" to select a new password. Complete help is available by selecting the bottom line "Help".
Characters you type are RED.
To have gcg initialized when you login:
- If you use the c-shell, add the following 2 lines to your .cshrc file:
source /seq/app/analysis_csh
gcg - If you use the k-shell, add the following 2 lines to your .kshrc file:
. /seq/app/analysis_ksh
gcg
Characters you type are RED.
To possibly correct the backspace key problem:
- If you use the c-shell, add the following line to your .cshrc file:
setenv TERM vt340
- If you use the k-shell, add the following line to your .kshrc file:
export TERM=vt340
The computers are available 24 hours a week, 7 days a week, with the exception of scheduled downtime, system crashes and critical maintenance.