The XYLEM Database Package

 XYLEM [Fristensky, (1993) NAR 21:5997-6003] is a package of tools designed to exploit the Unix environment to enable the user to identify, extract and manipulate data from major databases such as GenBank, EMBL and PIR. As shown above SPLITDB splits database files into annotation, sequence, and index files for more efficient searching.    Fundamental to the power of these programs is the ability to perform operations on groups of sequences, represented by names or accession numbers which function as virtual database subsets. Keyword searches can be performed by FINDKEY . Hits can be retrieved using FETCH . The most powerful program is FEATURES , which uses the GETOB parser to evaluate GenBank/EMBL/DDBJ Features Table expressions, thereby extract features (eg. mRNA, sig_peptide, intron) from lists of entries. Additional programs perform operations such as translation or randomization of datasets, and formatting of multiply-aligned sequences for publication. XYLEM is compatible with the Fristensky Sequence Analysis Package, and the Pearson FASTA programs [Pearson & Lipman, (1988) PNAS 85:2444-2448], and can be used from within the Genetic Data Environment (GDE) of Steven Smith [Smith et al. (1994) CABIOS 10:671-675].

Download compressed tar files:

Most recent version: XYLEM 1.8.7 (May 2002)

13 May 02 fil.py - Python script to create .fil files for use with fasta
10 May 02 a GDE menu item for splitdb has been added.
29 Jan 02 Updates have been made to gbupdate and pirupdate to
          conform to current directory organization at host sites.
25 Jan 02 GETLOC: Removed support for LiMB, Vecbase
          PIR files (-p) will be written with the .pir extension
          Previously, the .gen extension was used.

Source code and documentation ( xylem.1.8.7.tar.Z , 418k)
Solaris/Sparc binaries ( xylem.1.8.7.solaris-bin.tar.Z , 192k)
Linux/Intel binaries ( xylem.1.8.7.linux-bin.tar.Z , 179k)

XYLEM 1.8.6 (September 2001)

1. As of GenBank Release 127.0, the space allocated to LOCUS names is 16 characters, increased from a previous value of 12. GETOB and SPLITDB have been modified so that now both old and new formats can be read. In .ind files, the NAME filed will now occupy 16 characters.

2. Fixed bug in REFORM. Previously, any sequences over MAXLEN  in length would overflow the ALIGNMENT array. This isn't a problem if you directly compile the Pascal code, because the program would just crash, and the error would have been discovered.  Because I was usually translating to C, C tolerates  this kind of nonsense without a complaint. The bug has been fixed.

Source code and documentation ( xylem.1.8.6.tar.Z , 402k)
Solaris/Sparc binaries ( xylem.1.8.6.solaris-bin.tar.Z , 192k)
Linux/Intel binaries ( xylem.1.8.6.linux-bin.tar.Z , 169k)



XYLEM 1.8.3 BIRCH Return to BIRCH home page