[Beowulf] Updated NCBI rpms released for the 2.2.10 toolkit

Joe Landman landman at scalableinformatics.com
Thu Oct 28 14:06:09 EDT 2004


Folks:

  We rebuilt the NCBI rpms for AMD64, i386, i586 (non-P4), athlon, and 
i686 (p4).  Feel free to grab them from our site

   http://downloads.scalableinformatics.com/downloads/ncbi/

They are named NCBI-2.2.10-1.*.rpm, where * = 
{src,x86_64,i386,i586,i686}  They were built on RHEL/SuSE/Fedora Core2 
machines.  Should install without problems (and use the source if you 
have problems).

Please note that if you have a non-pentium4/non-athlon machine (PIII) 
you want the i586 or i386 version.  If you have a pentium4 based 
machine, you want the i686 version.  AMD64 (and probably EM64T) will use 
the x86_64.  Athlon's will use the athlon version.  Unless someone 
supplies me with G5 or Itanium2, I probably wont be able to do builds 
for those platforms.

Enjoy, and as usual, bug reports/problems to us, not to NCBI.  We built 
the RPMS, so if they are broken, we need to know.

Joe

-------- Original Message --------
Subject: 	[blast-announce] [ BLAST_Announce #044] BLAST 2.2.10 released
Date: 	Thu, 28 Oct 2004 11:35:37 -0400
From: 	Mcginnis, Scott (NIH/NLM/NCBI) <mcginnis at ncbi.nlm.nih.gov>
To: 	'blast-announce at ncbi.nlm.nih.gov' <blast-announce at ncbi.nlm.nih.gov>



Binaries can be obtained from:
ftp://ftp.ncbi.nlm.nih.gov/blast/executables/release/2.2.10/

Source code can be obtained from:
ftp://ftp.ncbi.nlm.nih.gov/toolbox/ncbi_tools/old/20041020/

Additionally, NCBI now provides anoncvs access
(http://www.ncbi.nlm.nih.gov/books/bv.fcgi?call=bv.View..ShowSection&rid=too
lkit.section.cvs_external) to toolkit sources. A cvsweb source browser
(http://www.ncbi.nlm.nih.gov/cvsweb/index.cgi/internal/c++/src/algo/blast/co
re/) and doxygen documentation
(http://www.ncbi.nlm.nih.gov/IEB/ToolBox/CPP_DOC/doxyhtml/group__AlgoBlast.h
tml) are also available.

Notes for the 2.2.10 release

New engine 

We have been rewriting and restructuring the BLAST engine in order to make
BLAST more modular and extensible. bl2seq and megablast currently support
the new engine; it can be enabled with the -V F option. Using the new engine
may result in significant performance improvements in some cases. 

General changes 

	-megablast now performs ungapped extensions in order to prevent
suboptimal alignments 
	-consolidated formatting code 
	-removed fmerge.c 
	-small fixes to sum statistics code 
	-better error handling 
	-fixed masking of translated queries 
	-fixed several readdb threading bugs 
	-improved protein neighbor generation 
	-hsp sorting/inclusion fixes 
	-many changes in HSP linking 
	-several fixes for translated RPS blast


BlastPGP 

	-added code to spread out gap costs when extracting data from
the sequence alignment to build PSSM 
	-changed handling of all-zero columns of residue frequencies to
use the underlying scoring matrix frequency ratios rather than
scoring matrix's scores 	
	- disallowed an ungapped search if more than one iteration is
specified 

scoremat.asn specification 

	-added a new 'formatrpsdb' application; given a collection of
Score-mat ASN.1 files, this application creates a database
suitable for use with RPSBLAST 
	-Simplified NCBI-ScoreMat specification to represent PSSMs
instead of arbitrary scoring matrices. blastpgp and formatrpsdb
can deal with this format.

If you have any questions please write to blast-help at ncbi.nlm.nih.gov


-- 
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web  : http://www.scalableinformatics.com
phone: +1 734 612 4615

_______________________________________________
Bioclusters maillist  -  Bioclusters at bioinformatics.org
https://bioinformatics.org/mailman/listinfo/bioclusters


-- 
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web  : http://www.scalableinformatics.com
phone: +1 734 612 4615

_______________________________________________
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf



More information about the Beowulf mailing list