readseq.c

EM算法的改进

/*
 * $Id: readseq.c 1339 2006-09-21 19:46:28Z tbailey $
 * 
 * $Log$
 * Revision 1.1  2005/07/29 17:18:51  nadya
 * Initial revision
 *
 */

/* File: readseq.c
 * main() program for ureadseq.c, ureadseq.h
 *
 * Reads and writes nucleic/protein sequence in various
 * formats. Data files may have multiple sequences.
 *
 * Copyright 1990 by d.g.gilbert
 * biology dept., indiana university, bloomington, in 47405
 * e-mail: [e-mail]
 *
 * This program may be freely copied and used by anyone.
 * Developers are encourged to incorporate parts in their
 * programs, rather than devise their own private sequence
 * format.
 *
 * This should compile and run with any ANSI C compiler.
 * Please advise me of any bugs, additions or corrections.
 *
 */

const char *title
    = "readSeq (1Feb93), multi-format molbio sequence reader.\n";

 /*  History
  27 Feb 90.  1st release to public.
   4 Mar 90.  + Gary Olsen format
              + case change
              * minor corrections to NBRF,EMBL,others
              * output 1 file per sequence for gcg, unknown
              * define -DNOSTR for c-libraries w/o strstr
              - readseq.p, pascal version, becomes out-of-date
  24 May 90.  + Phylip 3.2 output format (no input)
  20 Jul 90.  + Phylip 3.3 output (no input yet)
              + interactive output re-direction
              + verbose progress info
              * interactive help output
              * dropped line no.s on NBRF output
              * patched in HyperGCG XCMD corrections,
                - except for seq. documentation handling
              * dropped the IG special nuc codes, as IG has
                adopted the standard IUB codes (now if only
                everyone would adopt a standard format !)
  11 Oct 90.  * corrected bug in reading/writing of EMBL format

  17 Oct 91.  * corrected bug in reading Olsen format
                (serious-deletion)
  10 Nov 91.  * corrected bug in reading some GCG format files
                (serious-last line duplicated)
              + add format name parsing (-fgb, -ffasta, ...)
              + Phylip v3.4 output format (== v3.2, sequential)
              + add checksum output to all forms that have document
              + skip mail headers in seq file
              + add pipe for standard input == seq file (with -p)
              * fold in parts of MacApp Seq object
              * strengthen format detection
              * clarify program structure
              * remove fixed sequence size limit (now dynamic, sizeof memory)
              * check and fold in accumulated bug reports:
              *   Now ANSI-C fopen(..,"w") & check open failure
              *   Define -DFIXTOUPPER for nonANSI C libraries that mess
                  up toupper/tolower
              = No command-line changes; callers of readseq main() should be okay
              - ureadseq.h functions have changed; client programs need to note.
              + added Unix and VMS Make scripts, including validation tests

   4 May 92.  + added 32 bit CRC checksum as alternative to GCG 6.5bit checksum
                (-DBIGCHECKSUM)
    Aug 92    = fixed Olsen format input to handle files w/ more sequences,
                not to mess up when more than one seq has same identifier,
                and to convert number masks to symbols.
              = IG format fix to understand ^L

  25-30 Dec 92
              * revised command-line & interactive interface.  Suggested form is now
                  readseq infile -format=genbank -output=outfile -item=1,3,4 ...
                but remains compatible with prior commandlines:
                  readseq infile -f2 -ooutfile -i3 ...
              + added GCG MSF multi sequence file format
              + added PIR/CODATA format
              + added NCBI ASN.1 sequence file format
              + added Pretty, multi sequence pretty output (only)
              + added PAUP multi seq format
              + added degap option
              + added Gary Williams (GWW, [e-mail]) reverse-complement option.
              + added support for reading Phylip formats (interleave & sequential)
              * string fixes, dropped need for compiler flags NOSTR, FIXTOUPPER, NEEDSTRCASECMP
              * changed 32bit checksum to default, -DSMALLCHECKSUM for GCG version

   1Feb93
              = revert GenBank output to a fixed left number width which 
               other software depends on.
	      = fix for MSF input to handle symbols in names
	      = fix bug for possible memory overrun when truncating seqs for
		Phylip or Paup formats (thanks Anthony Persechini)

 */



/*
   Readseq has been tested with:
      Macintosh MPW C
      GNU gcc
      SGI cc
      VAX-VMS cc
   Any ANSI C compiler should be able to handle this.
   Old-style C compilers barf all over the source.


How do I build the readseq program if I have an Ansi C compiler?
#--------------------
# Unix ANSI C
# Use the supplied Makefile this way:
%  make CC=name-of-c-compiler
# OR do this...
% gcc readseq.c ureadseq.c -o readseq

#--------------------
$!VAX-VMS cc
$! Use the supplied Make.Com this way:
$  @make
$! OR, do this:
$ cc readseq, ureadseq
$ link readseq, ureadseq, sys$library:vaxcrtl/lib
$ readseq :== $ MyDisk:[myacct]readseq

#--------------------
# Macintosh Simple Input/Output Window application
# requires MPW-C and SIOW library (from APDA)
# also uses files macinit.c, macinit.r, readseqSIOW.make
#
Buildprogram readseqSIOW

#--------------------
#MPW-C v3 tool
C  ureadseq.c
C  readseq.c
link -w -o readseq -t MPST -c 'MPS '