the following awk script should do the job:

/^ID/   {printf(">%s;",$0); next;}
/^(PT|PA)/  {printf(" %s;",$0); next;}
/^\/\// {printf("\n"); next;}
/^    / {printf("\n%s",substr($0,5)); next;}
    {
    /* ignore default */
    }
END   {
    printf("\n");
    }

> awk -f file.awk file.txt
>ID  US74811111-0005    ; PT  I-003, a gene and methods for its use; PA  NIX CORPORATION RESEARCH TRIANGLE PARK, NC;
MDNNPNINECIPYNCLSNPEVEVLGGERIETGYTPIDISLSLTQFLLSEFVPGAGFVLGLVDIIWGIFGPSQWDAFPVQIEQLINQRIEEFARNQAISRLEGLSNLYVTIHEIENNTDELKFSNCVEEEIYPNNTVTCNDYTVNQEEYGGAYTSRNRGYNEAPSVPADYASVYEEKSYTDGRRENPCEFNRGYRDYTPLPVGYVTKELEYFPETDKVWIEIGETEGTFIVDSVELLLMEE

This Python script should work? Takes input file as an argument e.g. embllike2fasta.py infile.txt:

import sys
dict = {}
infile = open(sys.argv[1], "r")
outfile = open("outfile.fas", "w")
while 1:
    line = infile.readline()
    if not line:
        break
    parts = line.split(None, 1)
    if len(parts) == 1 and parts[0] == "//":
        outfile.write(">" + dict["ID"] + " ; " + dict["PT"] + " ; " + dict["PA"] + "\n")
        outfile.write(dict["seq"] + "\n")
    elif len(parts) == 1:
        dict["seq"] = parts[0].strip()
    else:
        dict[parts[0]] = parts[1].strip()
outfile.close()
infile.close()

I haven't tested it, and just hacked it up in the browser on my iPad, so it might not work? Plus it's late :-S lol!

Update: Tested it on my machine this morning, made a few quick edits! Very hacky, but it works :D

How about this? It's remarkably hacky, but should do the trick.

#!/usr/bin/env perl
use strict;
use warnings;
$/ = "//\n";
while (<>) {
    chomp;
    my %record;
    foreach my $line (split /\n/) {
        my ($key, $val) = unpack("A4A*", $line);
        $record{$key} .= $val;
    }
    print ">$record{ID} $record{PT} $record{PA}\n$record{''}\n";
}

The important bits are lines 4:($/ = "//n"), which sets the system to read one record at a time instead of a line, and 9:(unpack("A4A*", $line)), which splits the string into the EMBL key and the value.

I dont know for some reason this code is not working for this.. do you have any reason?? i apologize for this

ID   013789-0068
PS   TBD
OO   huringiensis
OS   ringiensis
OX
SI   68
RA
RL   2010. OKAYAMA UNIVERSITY,JAPAN LAMB CO LTD
FT   source          1..1176
MT
AC   67106
SV
CT
PN   013789
PT   PROTEIN PRODUCTION METHOD, FUSION PROTEIN, AND ANTISERUM
PA   AMA UNIVERSITY,JAPAN LAMB CO LTD.
PI   HAYAKAWA TORU (JP) SAKAI, HIROSHI, HAYAKAWA, TORU
P8
P4   10013789
P5   0
PC   International Classification: \nUS Classification: \nEuropean Classification: C12N15/62; C07K14/47A25
PR   80199166;
PE   199166
AN   09JP63603
KC   1
P1   ng the DNA into a host bacterium to transform the host bacterium; and (c) causing the expression of the fusion protein in the transformed host bacterium.; The method may further comprise a step of removing the peptide chain (B) from the fusion protein. \n \n
P7
P9   112
PO
PM   10013789;
PB   10013789
PQ   10013789;
EM   esentative
W1   PRT
D1   0204
D2   0217
D3   0730
D4   0801
D5   0204
HL   [L[P9_GQ;0;3,WO2010013789,45,67]] [L[PM_PN_GQNUC;0;12,WO2010013789]] [L[PQ_PN_GQNUC;0;12,WO2010013789]]
CC   mer C1-1-f FH   Key             Location/Qualifiers Copyright (c)Inc. 2011
LS   Application
L2   Publ. Of int. appl. w4

  MDNNPNINECIPYNCLSNPEVEVLGGERIETGYTPIDISLSLTQFLLSEFVPGAGFVLGLVDIIWGIFGPSQWDAFPVQI
  EQLINQRIEEFARNQAISRLEGLSNLYQIYAESFREWEADPTNPALREEMRIQFNDMNSALTTAIPLLAVQNYQVPLLSV
  YVQAANLHLSVLRDVSVFGQRWGFDAATINSRYNDLTRLIGNYTDYAVRWYNTGLERVWGPDSRDWVRYNQFRRELTLTV
  LDIVALFSNYDSRRYPIRTVSQLTREIYTNPVLENFDGSFRGMAQRIEQNIRQPHLMDILNSITIYTDVHRGFNYWSGHQ
  ITASPVGFSGPEFAFPLFGNAGNAAPPVLVSLTGLGIFRTLSSPLYRRIILGSGPNNQELFVLDGTEFSFASLTTNLPST
  IYRQRGTVDSLDVIPPQDNSVPPRAGFSHRLSHVTMLSQAAGAVYTLRAPTFSWQHRSAEFNNIIPSSQITQIPLTKSTN
  LGSGTSVVKGPGFTGGDILRRTSPGQISTLRVNITAPLSQRYRVRIRYASTTNLQFHTSIDGRPINQGNFSATMSSGSNL
  QSGSFRTVGFTTPFNFSNGSSVFTLSAHVFNSGNEVYIDRIEFVPAEVTFEAEYDLERAQKAVNELFTSSNQIGLKTDVT
  DYHIDQVSNLVECLSDEFCLDEKQELSEKVKHAKRLSDERNLLQDPNFRGINRQLDRGWRGSTDITIQGGDDVFKENYVT
  LLGTFDECYPTYLYQKIDESKLKAYTRYQLRGYIEDSQDLEIYLIRYNAKHETVNVPGTGSLWPLSAQSPIGKCGEPNRC
  APHLEWNPDLDCSCRDGEKCAHHSHHFSLDIDVGCTDLNEDLGVWVIFKIKTQDGHARLGNLEFLEEKPLVGEALARVKR

Or just:

egrep -e '^(ID|  )' | sed -e 's/^ID  />/g' -e 's/^    //g'

Oh, I see you wanted the contents of PT there too. I leave it as an excercise for the reader! :-)

This one-liner should do it:

perl -ne 'if ($_ =~ /^ID  (.+)/) { print ">$1" } elsif ($_ =~ /^P[TA]  (.+)/) { print "; $1 " } elsif ($_ =~ /^\s{4}(.+)/) { print "\n$1\n" }'

Output:

>US74811111-0005    ; I-003, a gene and methods for its use ; NIX CORPORATION RESEARCH TRIANGLE PARK, NC 
MDNNPNINECIPYNCLSNPEVEVLGGERIETGYTPIDISLSLTQFLLSEFVPGAGFVLGLVDIIWGIFGPSQWDAFPVQIEQLINQRIEEFARNQAISRLEGLSNLYVTIHEIENNTDELKFSNCVEEEIYPNNTVTCNDYTVNQEEYGGAYTSRNRGYNEAPSVPADYASVYEEKSYTDGRRENPCEFNRGYRDYTPLPVGYVTKELEYFPETDKVWIEIGETEGTFIVDSVELLLMEE