Previous - Next - Bottom - PP home - PP help TOC

PP Help 10: News

NEWS

PP is migrating...

Dear user,

   PredictProtein  has moved from EMBL (Heidelberg) to CUBIC (New York)!

   I am glad to acknowledge the  support from various 'sources' enabling
the move to Columbia University:
* David Hirsh  (Dep of Biochem & Mol Biophysics, Columbia) unearthed the
  start-up money required for buying a machine to run PP from University
  sources.
* Julie Nash (Silicon Graphics)  successfully worked out the  conditions
  for a 'special deal' to purchase an Origin 2000 on which PP is running
  now.
* Further support from Columbia University came from:
  Barry Honig and Kathy Rosa (Biochem & Mol Biophys) contributed import-
  antly to the fact that PP is already running on a machine in New York.
  Volker Eyrich (Chemistry) and  Murad Nayal (Biochem & Mol Biophys) as-
  sisted in appropriately setting up the machine running PP.
  
   PredictProtein is running on 8 processors of an SGI Origin 2000:
  	dodo.cpmc.columbia.edu.

   One problem that may  immediately become apparent is the insufficient
internet connection of Columbia University.   The major problem seems to
be caused by Columbia's internet provider:  Applied Theory, respectively
Sprint.  We are working on this problem.

   Finally,  please note that some default settings have changed,  and a
few new options - in particular for advanced users - have been implemen-
ted.
							   Burkhard Rost
				     Columbia Univ, New York, March 1999

Currently known problems

1. none

PP WWW pages are mirrored to ease access times

1. 09.98
   Australia
   Site: ANGIS centre, Sydney University
   Contact: Shoba Ranganathan (shoba@angis.org.au)
2. BIC, Singapore
   Site: BioInformatics Centre, National Univ. of Singapore
   Contact: Betty Chenk (betty@bic.nus.edu.sg)
3. 03.99
   CUBIC New York
   Site: Columbia University BioInformatics Centre
   Contact: Burkhard Rost (rost@parrot.bioc.columbia.edu)
4. 04.99
   CNB Madrid (Spain)
   Site: National Centre of Biotechnology
   Contact: Jose Valverde (jrvalverde@cnb.uam.es)
5. 05.99
   CDFD Hyderabad (India)
   Site: Centre for DNA Fingerprinting and Diagnostics (CDFD)
   Contact: Kunchur Gururprasad (guru@salarjung.embnet.org.in )
6. 07.99
   Univ Rome
   Site: University 'Tor Vergata'
   Contact: Manuela Helmer-Citterich (citterich@uniroma2.it
7. 07.99
   Univ of Pune (India)
   Site: Bioinformatics Centre, University of Pune
   Contact: Sunita Jagtap (sunita@bioinfo.ernet.in)

8. Who wants to be next?

PP news 1999

META PredictProtein

• Miscellaneous services
  1. SignalP prediction of presence and location of signal peptide cleavage sites in amino acid sequences from different organisms
  2. NetOglyc prediction of mucin type GalNAc O-glycosylation sites in mammalian proteins.
  3. NetPicoRNA predictions of cleavage sites of picornaviral proteases.
  4. ChloroP predictions of whether or not a protein contains an N-terminal chloroplast transit peptide, cTP, and of probable sites for cleavage of the transit peptide.
• Secondary structure prediction
  1. JPRED consensus method for protein secondary structure prediction.
• Transmembrane helices
  1. TMHMM prediction the location of transmembrane helices and their topology.
  2. TOPPRED prediction of location and orientation of transmembrane helices.
  3. DAS prediction of location of transmembrane helices.
• Threading (remote homology search)
  1. FRSVR prediction-based threading, also incorporating purely sequence-based database searches.
  2. SAMT98 hidden Markov model method (SAM-T98) for finding remote homologs of protein sequences.
• Homology modelling
  1. SWISS-MODEL prediction of 3D structure by homology modelling (automated server).
  2. CPHmodels prediction of 3D structure by homology modelling through a collection of methods and databases developed to predict protein structures.

New prediction defaults

1. 04.99
   Low-complexity regions by SEG (default, example): The program SEG (J C Wootton & S Federhen) is executed, by default. SEG scans your sequence for regions of low-complexity ('simple sequences' or 'composition-biased regions'). You may turn this default off by using the keyword 'no seg in any line before the one beginning with a hash ('#').
2. 03.99
   Filtering your alignment (default): Your alignment will be filtered before running PHD, in order to reduce possible redundancy. You may turn this default off by using the keyword 'no filter.

New output option

1. 04.98
   Return no seg ('return no seg')
   The string "return no seg" in any line before the one starting with a hash (#) results in that SEG results are not returned.
2. 04.99
   Store results here, return no mail output ('return no mail')
   The string "return no mail" in any line before the one starting with a hash (#) results in that we shall not returned the results by mail. Instead, the results for your requests will be stored on our machines for 3 days, and you will receive a mail that simply tells you how you can ftp the result from here. The reason for including this option is that some requests may result in very large output files, and those may be difficult to handle for your local mailing device (in particular when you request HTML formatted output).
3. 04.99
   Return output in HTML format ('return html')
   The string "return html" in any line before the one starting with a hash (#) results in that the email you get will have the entire results attached in one HTML formatted file (which you can load into any WWW browser).
   The conversion of the old format to the new one is still rather rudimentary. However, thanks to the program MView (Nigel Brown, MRC, Mill Hill, London) part of the output is already much easier to digest in this new format!
4. 04.99
   Return output in HTML format for printouts ('perline=60')
   The strings "return html60" OR "perline=N" (with N=1..number of residues of your protein) in any line before the one starting with a hash (#) result in that the email you get will have the all output attached in one HTML formatted file (to display with any WWW browser) that has fewer characters per line than the normal HTML output (see "return HTML"), so that you can print the output. (For further information see the option "return html".)
5. 04.99
   Return output in HTML format (with PHD graphics) ('return html detail')
   The string "return html detail" in any line before the one starting with a hash (#) results in that the email you get will have the entire results attached in one HTML formatted file (which you can load into any WWW browser), furthermore, the PHD predictions will also be displayed graphically.
   Note: the HTML files resulting from the PHD predictions may be large. To avoid that your mail will be too big, you may therefore use the option of leaving the result on our machines, and simply ftp it to your local machine (see option "return no mail").
6. 04.99
   Return output (with PHD graphics) in HTML format for printouts ('perline=60')
   The strings "return html detail 60" OR "perline=N" (with N=1..number of residues of your protein) in any line before the one starting with a hash (#) result in that the email you get will have the all output attached in one HTML formatted file (to display with any WWW browser) that has fewer characters per line than the normal HTML output (see "return HTML detail"), so that you can print the output. (For further information see the option "return html".)
   Note: the HTML files resulting from the PHD predictions may be large. To avoid that your mail will be too big, you may therefore use the option of leaving the result on our machines, and simply ftp it to your local machine (see option "return no mail").

New input option

1. 03.99
   Your input sequence by its SWISSPROT identifier ('# swissid')
   You can submit your sequence through its SWISS-PROT identifier. The string "# SWISSID" prompts the system to interpret the following line to be an identifier of the form 'name_species' (example for input).
   NOTE: only identifiers in our current SWISS-PROT release are valid.
2. 03.99
   Your input sequence(s) by FASTA alignment ('do not align')
   You can submit your alignment in FASTA format. This requires two keywords: (1) the string "# FASTA list" prompts the system to interpret your alignment as one in the FASTA format (specification and example for FASTA list format), and (2) the keyword "do NOT align" in a line before the one starting with a hash ('#'), assures that your alignment will not be re-aligned (as for the options of submitting your sequences via FASTA format).
   
3. 03.99
   Your input sequence(s) by PIR alignment ('do not align')
   You can submit your alignment in PIR format. This requires two keywords: (1) the string "# PIR list" prompts the system to interpret your alignment as one in the PIR format (specification and example for PIR list format), and (2) the keyword "do NOT align" in a line before the one starting with a hash ('#'), assures that your alignment will not be re-aligned (as for the options of submitting your sequences via PIR format).
   

PP news 1998

1. 02.98
   Your input sequence(s) in FASTA-list format ('# FASTA list ')
   You can submit your alignment in a format which is hopefully easier for you to write than the MSF format. The string "# FASTA list" prompts the system to interpret your alignment as one in the SAF format (specification and example for FASTA list format).
   

New prediction defaults

1. 04.98
   Coiled-coil regions: now by default the program COILS written by Andrei Lupas is run on your sequence. An output is returned if a coiled-coil region has been detected.
2. 04.98
   Functional sequence motifs: now by default the PROSITE database written by Amos Bairoch, Philip Bucher and Kay Hofmann is scanned for sequence motifs. An output is returned if any motif has been detected.

New output options

1. 04.98
   Return also BLASTP output ('return blast')
   The string "return blast" in any line before the one starting with a hash (#) returns additionally the unfiltered BLASTP output.
2. 04.98
   Return prediction additionally in RDB format ('return phd rdb')
   The string "return phd rdb" in any line before the one starting with a hash (#) returns the PHD predictions additionally in RDB format (as read and written by local versions of the programs PHD and TOPITS).
3. 04.98
   Return topits hssp ('return topits hssp')
   The string "return topits hssp" in any line before the one starting with a hash (#) returns the threading output additionally in HSSP format.
4. 04.98
   Return no coils ('return no coils')
   The string "return no coils" in any line before the one starting with a hash (#) results in that COILS results are not returned.
5. 04.98
   Return no prosite ('return no prosite')
   The string "return no prosite" in any line before the one starting with a hash (#) results in that PROSITE results are not returned.

PP news 1997

1. 06.97
   Your input alignment in SAF format ('# SAF format ')
   You can submit your alignment in a format which is hopefully easier for you to write than the MSF format. The string "# SAF" prompts the system to interpret your alignment as one in the SAF format (specification and example for SAF format).
   

New output options

1. 03.97
   Return prediction additionally in MSF format ('return phd msf')
   The string "return phd msf" in any line before the one starting with a hash (#) returns the PHD predictions additionally in an MSF format (appended to the alignment).
   
2. 03.97
   Return prediction additionally in CASP2 format ('return phd casp2')
   The string "return phd casp2" in any line before the one starting with a hash (#) returns the PHD predictions additionally in the format used for the second protein structure prediction contest in Asilomar, 1996 (CASP2).
   

PP news 1996

Notes

1. The error rate in falsely predicting transmembrane helices for globular proteins has been reduced by a new program to below 2%.

New services

1. 11.96: Estimate for approximate response time of PredictProtein
2. 03.96: Refined prediction of transmembrane helices and topology (PhdTopology)

• Refined prediction of transmembrane helices and topology (help on topology prediction; examples for: request; and output)

  The neural network prediction of transmembrane helices (PHDhtm) is refined by a dynamic programming-like algorithm. This method resulted in correct predictions of all transmembrane helices for 89% of the 131 proteins used in a cross-validation test; more than 98% of the transmembrane helices were correctly predicted. The output of this method is used to predict topology, i.e., the orientation of the N-term with respect to the membrane. The expected accuracy of the topology prediction is > 86%. Prediction accuracy is higher than average for eukaryotic proteins and lower than average for prokaryotes. PhdTopology is more accurate than all other methods tested on identical data sets (Rost, Casadio & Fariselli, submitted, 1996).

1. 03.96: For predicting the locations of transmembrane helices, PHDhtm is used. The more accurate refined version along with the prediction of topology (PHDtopology) is available upon request ('predict htm topology').

Previous - Next - Top - PP home - PP help TOC