ProtoGene.pl 57.9 KB
Newer Older
Sebastien Moretti's avatar
Sebastien Moretti committed
1
2
3
4
5
6
7
8
9
10
11
12
#!/usr/bin/env perl

########################################################
#                                                      #
# ProtoGene : Bona-Fide Back Translation of Protein    #
#             Multiple Sequence Alignments             #
#                                                      #
########################################################

use strict;
use warnings;
use diagnostics;
Sebastien Moretti's avatar
Sebastien Moretti committed
13

14
15
16
17
18
use File::Which qw(which);                 # Locate external executable programs in the PATH
use Time::localtime;                       # Use localtime+PID for a pseudo-uniq temp file name
use Getopt::Long;                          # Options specifications
use File::Copy qw(move);                   # Avoid external 'mv' command usage
use LWP::Simple;                           # To test gigablaster availability
Sebastien Moretti's avatar
Sebastien Moretti committed
19

20
use Mail::Send;                            # Send warnings and errors files by e-mail ==> only if the $userEMail variable is defined
Sebastien Moretti's avatar
Sebastien Moretti committed
21

22
23
# TODO use FinBin for lib and binary paths !!!
use lib '/software/SequenceAnalysis/ProtoGene/4.2.0/lib/'; # Local path for ProtoGene's own perl modules
24
25
26
use Exonerate;                             # Exonerate runner, parser, ...
use Views;                                 # Non-text outputs, e.g. HTML/CSS
#use CheckOutput;                         # Check output for cds consistancy with query
Sebastien Moretti's avatar
Sebastien Moretti committed
27
28
29



30
################## CONFIGURATION ##################
31
my $cachePath        = '/scratch/beegfs/monthly/tcoffee/ProtoGene_Cache';   # Cache directory
32
33
34
my $cacheStorageTime = 15;                                                  # Do not update sequences younger than X days

my $userEMail        = 'moretti.sebastien@gmail.com';                       # To receive e-mails with encountered problems; leave blank to inactive
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

### BLAST parameters ###
my $blast_param = { 'evalue'    => 0.05,
                    'matrix'    => 'BLOSUM62',
                    'filter'    => 'Off',
                    'species'   => 'All_organisms',

                    'db1'       => 'refseq_protein', # NCBI RefSeq protein
                    'identity1' => 100,
                    'cover1'    => 100,
                    'db2'       => 'nr', # NCBI nr protein
                    'identity2' =>  95,
                    'cover2'    =>  95,

                    'nbesthits' => 10,
50
                    'core'      => 1, # TODO really used ?
51
                  };
52
###################################################
Sebastien Moretti's avatar
Sebastien Moretti committed
53
54


55
my $VERSION  = '4.2.2';
Sebastien Moretti's avatar
Sebastien Moretti committed
56

57
my $webblast_exe  = '/software/SequenceAnalysis/ProtoGene/4.2.0/bin/webblast.pl';
58
59
my $blast_exe     = 'blastall';        # Or wu-blastall for Wu-BLAST; for local blast usage
my $exonerate_exe = 'exonerate';       # Exonerate 1.0 because current parser only works with this version
60
61
62


################## Option management
63
my ($msa, $revtrans, $pep, $hideBOJ, $run_name, $template, $lim) = ('', 0, 0, 0, '', '', 0);
64
my ($cache, $cleancache)                                         = ('update', 'update');#TODO Finish to implement
65
66
my ($debug, $tmp)                                                = (0, 0);
my ($db, $species, $local, $giga)                                = ($blast_param->{'db1'}, $blast_param->{'species'}, 0, 0);
Sebastien Moretti's avatar
Sebastien Moretti committed
67
my %opts = ('msa|in=s'        => \$msa,        # Input sequences
68
            'revtrans:s'      => \$revtrans,   # Use to reverse-translate sequences with no match
Sebastien Moretti's avatar
Sebastien Moretti committed
69
70
71
72
73
            'pep'             => \$pep,        # Add the original peptide query beneath the related CDS seq
            'hideBOJ'         => \$hideBOJ,    # Hide BOJ output
            'run_name=s'      => \$run_name,   # Use another name, instead of input seq name, for result files
            'template=s'      => \$template,   # Use a template file
            'lim=i'           => \$lim,        # Limit number of input query sequences
74
75
            'cachedir=s'      => \$cache,      # Cache directory
            'cacheclean=s'    => \$cleancache, # Cache behavior
Sebastien Moretti's avatar
Sebastien Moretti committed
76
77
78

            'orgm|species=s'  => \$species,    # Organism(s) to blast against
            'db|database=s'   => \$db,         # Database to blast against
79
            'local'           => \$local,      # Use to specify a local db query, defined in $local_db
Sebastien Moretti's avatar
Sebastien Moretti committed
80
81
82
83
84
85
86
87
88
89
            'giga'            => \$giga,       # Use GigaBlaster server

            'version|V'       => sub { print "\n\tPROTOGENE : Bona-Fide Back Translation of Protein Multiple Sequence Alignments\n\tversion   : $VERSION\n\n";
                                       exit 0;
                                 },            # Print version information
            'help|H|?'        => sub { system('perldoc', "$0")==0 || print {*STDERR} "\n\tCannot open the documentation with 'perldoc'\n\n";
                                       exit 0;
                                     },        # Print full help message
            'debug'           => \$debug,      # Verbose output
            'tmp'             => \$tmp,        # To keep traces of fake intermediate files like fake xml from NCBI, fake aln, ...
Sebastien Moretti's avatar
Sebastien Moretti committed
90
           );
Sebastien Moretti's avatar
Sebastien Moretti committed
91
my $test_option_values = Getopt::Long::GetOptions(%opts);
92
#$revtrans = 1  if ( $revtrans eq '' ); # Allow revtrans to be a boolean or a string option (for tcoffee web server)
Sebastien Moretti's avatar
Sebastien Moretti committed
93

94
95

################## Short help message
96
if ( !$test_option_values || ($msa eq '' && $cleancache ne 'empty' && $cleancache ne 'old') ){
97
98
99
    print {*STDERR} "\n\tCannot open the MSA file in FASTA format
\tTry:  $0 --msa=path_of_the_fasta_msa_file [Options]

100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
\tOptions: --orgm      =All_organisms, Bacteria, Viruses, Vertebrata,
\t                      Eukaryota, Mammalia, Primates, Homo_sapiens,
\t                      Gallus_gallus, Bos_taurus, Escherichia_coli,
\t                      Arabidopsis_thaliana, Mus_musculus,
\t                      Drosophila_Melanogaster, ...
\t                      default is '$blast_param->{'species'}'
\t         --db        =nr, pdb, swissprot, refseq_protein
\t                      default is '$blast_param->{'db1'}'
\t         --local      to execute a local BLAST query with
\t                      --db=path_for_a_local_db_blast_formated\n
\t         --template   to provide your own nucleotidic sequences
\t                      following the cds file format
\t         --revtrans   reverse-translates sequences with no
\t                      blast hit, in IUB (IUPAC) depiction code
\t                      They are removed from the alignement by default
\t         --pep        adds the original peptide query beneath the
\t                      back-translated sequence
\t         --cachedir  ='own_PATH_directory'
\t                      (default is '$cachePath')
\t         --cacheclean=none, update, use, old, empty
\t                      to select the cache behavior (default is 'update')\n
\t         --debug      prints extra information when running
\t         --version    prints version information
\t         --help       prints a full help message\n\n";
124
    exit(1);
Sebastien Moretti's avatar
Sebastien Moretti committed
125
}
126
127


128
################## Check cache directory accessibility
Sebastien Moretti's avatar
Sebastien Moretti committed
129
130
131
if ( -d $cache ){
    $cachePath = $cache;
}
132
checkCacheAccessibility($cachePath);
133

134
135
# Cache management
$cache = cacheManagement($cache, $cachePath, 0) || '.';
136

137
138
# Remove old files from the cache directory everytime ProtoGene is running
cacheManagement('old', $cachePath, 1);
139
140


141
142

################## Check external software in the PATH
143
checkExternalSoftware($webblast_exe, $exonerate_exe);
144
145
if ( $local ){
    checkExternalSoftware($blast_exe);
Sebastien Moretti's avatar
Sebastien Moretti committed
146
147
148
149
}



150
151
152
################## Check input files
# Open and check template file
my ($template_NT, $template_AA, $template_SEQ);
153
#TODO: template is fasta only now
154
155
($template_NT, $template_AA, $template_SEQ) = checkTemplate($template) if ( $template );

Sebastien Moretti's avatar
Sebastien Moretti committed
156

Sebastien Moretti's avatar
Sebastien Moretti committed
157
158
159
160
161
# Open and Check the input fasta file
my $originalMSA    = $msa;
my $converted      = 0; # Accept only fasta format
#TODO: Do a better file converter and identifier is msa is not fasta
#($msa, $converted) = isFastaFile($msa, $converted);
Sebastien Moretti's avatar
Sebastien Moretti committed
162

163
my $fasta_checker = -1;
164
my (@original_names, @original_seq); #Use 2 lists in parallel : original_names, seq
Sebastien Moretti's avatar
Sebastien Moretti committed
165
166
open( my $MSA, '<', "$msa" ) or die "\n\tCannot open your MSA file\n\n";
READ_MSA:
167
while(<$MSA>){
Sebastien Moretti's avatar
Sebastien Moretti committed
168
169
    next READ_MSA if ( $_ =~ /^#/ );

170
    if ( $_ =~ /^>/ ){
171
172
        $fasta_checker++;
        my $name = $_;
Sebastien Moretti's avatar
Sebastien Moretti committed
173
174
175
        $name    =~ s{\r\n}{}g; #Remove return lines from windows OS '^M'
        $name    =~ s{\r}{\n}g;
        $name    =~ s{^>[ \t]+}{>};
176
        chomp($name);
Sebastien Moretti's avatar
Sebastien Moretti committed
177
178

        push @original_names, $name;
Sebastien Moretti's avatar
Sebastien Moretti committed
179
    }
Sebastien Moretti's avatar
Sebastien Moretti committed
180
181
182
183
184
185
186
187
188
189
    elsif ( $_ =~ /[A-Z\-\.]/i ){
        my $sequence = $_;
        $sequence    =~ s{\r\n}{}g;
        $sequence    =~ s{\r}{\n}g;
        $sequence    =~ s{\.}{-}g; #for msa with '.' as gap
        #TODO because exonerate cannot work with selenocystein
        $sequence    =~ s{[BJOUZ]}{X}ig; #U here for selenocystein
        $sequence    =~ s{[^A-Za-z\-\*\n]}{}g; #Remove all the non-gap or non-alphabetic characters from the seq
        chomp($sequence);

Sebastien Moretti's avatar
Sebastien Moretti committed
190
        #fasta sequence on 1 line
191
        if ( !exists($original_seq[$fasta_checker]) ){
Sebastien Moretti's avatar
Sebastien Moretti committed
192
            push @original_seq, $sequence;
193
194
        }
        else {
Sebastien Moretti's avatar
Sebastien Moretti committed
195
            $original_seq[$fasta_checker] .= $sequence;
196
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
197
198
    }
}
199
close $MSA;
Sebastien Moretti's avatar
Sebastien Moretti committed
200
201

#Are there FASTA sequences ?
202
if ( $fasta_checker == -1 ){
Sebastien Moretti's avatar
Sebastien Moretti committed
203
    failure();
Sebastien Moretti's avatar
Sebastien Moretti committed
204
205
    print {*STDERR} "\tThe MSA file does NOT seem to be a protein FASTA format
\tSee an example of FASTA format <a href='http://en.wikipedia.org/wiki/FASTA_format' target='_blank'>here</a>\n\n";
206
    exit(1);
Sebastien Moretti's avatar
Sebastien Moretti committed
207
}
208
elsif ( $lim >0 && $fasta_checker > $lim ){
Sebastien Moretti's avatar
Sebastien Moretti committed
209
    failure();
210
    print {*STDERR} "\tThe FASTA file is too large, try with less than $lim sequences\n\tor split your file\n\n";
211
212
    exit(1);
}
213
elsif ( exists( $original_seq[0] ) && $original_seq[0] =~ /[acgtu]/i ){
214
    my $first_seq = $original_seq[0];
Sebastien Moretti's avatar
Sebastien Moretti committed
215
216
217
218
219
220
    #Check if sequences are amino acids and not nucleotides ONLY on the 1st sequence
    my $a         = ($first_seq =~ s{[aA]}{}g)               || 0;
    my $c         = ($first_seq =~ s{[cC]}{}g)               || 0;
    my $g         = ($first_seq =~ s{[gG]}{}g)               || 0;
    my $t         = ($first_seq =~ s{[tTuU]}{}g)             || 0;
    my $non       = ($first_seq =~ s{[^aAcCgGtTuUXxNn-]}{}g) || 0;
221
    if ( ($a+$c+$g+$t) >= (($a+$c+$g+$t+$non)*80/100) ){
Sebastien Moretti's avatar
Sebastien Moretti committed
222
        failure();
Sebastien Moretti's avatar
Sebastien Moretti committed
223
224
        print {*STDERR} "\tYour sequences look already to be nucleotides
\tThe goal of this program is to turn AMINO ACID alignments into CDS nucleotide alignments\n\n";
225
226
        exit(1);
    }
Sebastien Moretti's avatar
Sebastien Moretti committed
227
228
229
230
}
undef $fasta_checker;


Sebastien Moretti's avatar
Sebastien Moretti committed
231
# Check GigaBlaster status if used
232
if ( $giga==1 ){
233
    $local = 0;
Sebastien Moretti's avatar
Sebastien Moretti committed
234
    $local = 2  if ( head('http://www.igs.cnrs-mrs.fr/public_g/remoteblast.cgi') );
235
236
}

Sebastien Moretti's avatar
Sebastien Moretti committed
237

Sebastien Moretti's avatar
Sebastien Moretti committed
238
# Temporary file extension
239
my $date = sprintf("%d-%02d-%02d_%02dh%02d", localtime->year() + 1900, localtime->mon() + 1, localtime->mday(), localtime->hour, localtime->min);
Sebastien Moretti's avatar
Sebastien Moretti committed
240
241


242

Sebastien Moretti's avatar
Sebastien Moretti committed
243
#Start main program with version # of programs and list original queries
244
print "\n\t   Protogene\t$VERSION\t$date\n\n";
245

246
open(my $EXONERATEISHERE, "$exonerate_exe --version |");
247
my $IsExonerateHere = <$EXONERATEISHERE>;
248
close $EXONERATEISHERE;
Sebastien Moretti's avatar
Sebastien Moretti committed
249
250
print "\t   $IsExonerateHere\n";
undef $IsExonerateHere;
251
undef $VERSION;
252

Sebastien Moretti's avatar
Sebastien Moretti committed
253
for(my $m=0; $m<=$#original_names; $m++){
254
    printf("\t>%-3s  is  %s\n", $m, $original_names[$m]);
Sebastien Moretti's avatar
Sebastien Moretti committed
255
}
Sebastien Moretti's avatar
Sebastien Moretti committed
256
$date .= "_$$"; #Add PID to be more uniq
Sebastien Moretti's avatar
Sebastien Moretti committed
257
258
259
260



# run_name implementation for web server
261
$originalMSA = $run_name  if ( $run_name ne '' );
Sebastien Moretti's avatar
Sebastien Moretti committed
262

263
264
unlink("$originalMSA.cds", "$originalMSA.cdsP", "$originalMSA.cdsP.html",
       "$originalMSA.out", "$originalMSA.boj",  "$originalMSA.log");
Sebastien Moretti's avatar
Sebastien Moretti committed
265
266


267
# Build the sequences, from the alignment, to perform blast search
268
EACH_SEQ:
269
270
271
272
273
274
for(my $r=0; $r<=$#original_names; $r++){
    my $fasta_header = $original_names[$r];
    my $short_name   = $original_names[$r];
    $fasta_header    =~ s{^>}{};               # Full FASTA header
    $short_name      =~ s{^([^\s\t]+).*$}{$1}; # Short name without description

Sebastien Moretti's avatar
Sebastien Moretti committed
275

Sebastien Moretti's avatar
Sebastien Moretti committed
276
    my @equivalent_blast_hits; #Multiple BLAST hits with the same highest e-value/score
277
278

    if ( exists($template_NT->{$short_name}) && $template_NT->{$short_name} ne '' ){
Sebastien Moretti's avatar
Sebastien Moretti committed
279
280
281
        # First try nucleotide template
        push @equivalent_blast_hits, 'My_own_seq';
        print "\n\n$r done\n\nNucleotide template for $r:\t$template_NT->{$short_name}\n\n*******************************************************************\n\n";
Sebastien Moretti's avatar
Sebastien Moretti committed
282
    }
283
    elsif ( exists($template_AA->{$short_name}) && $template_AA->{$short_name} ne '' ){
Sebastien Moretti's avatar
Sebastien Moretti committed
284
285
286
        # Second try amino acid/BLASTp template
        push @equivalent_blast_hits, $template_AA->{$short_name};
        print "\n\n$r done\n\nBlastP template for $r:\t$template_AA->{$short_name}\n\n*******************************************************************\n\n";
Sebastien Moretti's avatar
Sebastien Moretti committed
287
288
    }
    else{
289
    # else send them to webblast
Sebastien Moretti's avatar
Sebastien Moretti committed
290
291
292
293
294
295
296
        # Prepare sequence $r for BLAST
        open(my $READY2BLAST, '>', "$cache/$date.seq2blast.fas.$r");
        my $noGap = $original_seq[$r];
        $noGap    =~ s{\-}{}g; #remove gaps
        print {$READY2BLAST} ">$r\n$noGap\n";
        close $READY2BLAST;

297
298
299
300
301
        if ($r==0){
            # Show BLAST parameters only for the first BLAST run
            display_blast_param();
        }
        run_BLAST("$cache/$date.seq2blast.fas.$r", $local, $blast_param->{'db1'}, $blast_exe, "$cache/$date.blastp.$r");
Sebastien Moretti's avatar
Sebastien Moretti committed
302

303
304
305

        # Get the BLASTp hit(s)  acc number
        open(my $BLASTP, '<', "$cache/$date.blastp.$r");
306
        while(<$BLASTP>){
307
            push @equivalent_blast_hits, $1  if ( $_ =~ /^>.+?\@.+?__(\S+).*$/ ); #Warning: double '_'
308
309
        }
        close $BLASTP;
310
        unlink("$cache/$date.blastp.$r", "$cache/$date.seq2blast.fas.$r")  if ( $tmp == 0 || exists($equivalent_blast_hits[0]) );
Sebastien Moretti's avatar
Sebastien Moretti committed
311
        if ( !exists($equivalent_blast_hits[0]) ){
312
            print "\n$r ...\tNo blast result found above thresholds for $fasta_header\n\n";
313
            buildFailureOutputFiles($r, 'No_BLASTp_Result', 'Not_searched');
314
315
316
317
            undef $original_seq[$r];
            undef $original_names[$r];
            next EACH_SEQ;
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
318
319
320
321
    }



322
    #When multiple equivalent blast hits  ==> use, and add, every hits
323
    my ($resultPOS, $resultBOJ) = ('', '');
324
    my @failureStatus = (''); # To always have $failureStatus[0] defined
325
    my @intronStatus  = ('');
326
    HIT_LINK:
Sebastien Moretti's avatar
Sebastien Moretti committed
327
328
    for(my $qq=0; $qq<=$#equivalent_blast_hits; $qq++){
        print "\n$fasta_header  -->  [$equivalent_blast_hits[$qq]]\n";
329
330
331
332


        my @nt_GIs;
        my $intronStep = 0;
333
334
335
        # Get Nt sequence from template
        if ( exists($template_NT->{$short_name}) && $template_NT->{$short_name} ne '' ){
            if ( $template_NT->{$short_name} !~ /^My_Seq$/i ){
336
                download_seq($cache, '', '', $template_NT->{$short_name});
337
338
339
340
341
342
343
344
                @nt_GIs = $template_NT->{$short_name};
            }
            else {
                open(my $MYSEQ, '>', "$cache/My_Seq-$$.fas");
                print {$MYSEQ} ">My_Seq-$$ for $short_name\n".$template_SEQ->{$short_name}."\n";
                close $MYSEQ;
                @nt_GIs = "My_Seq-$$";
            }
345
        }
346
        # Get Nt sequence through prot acc -> prot gi -> nt linked gi -> nt acc
347
        else{
Sebastien Moretti's avatar
Sebastien Moretti committed
348
            # Prot ACC -> PUID
Sebastien Moretti's avatar
Sebastien Moretti committed
349
            my $protGI = blastPAcc2PGI($equivalent_blast_hits[$qq]);
350
            if ( $protGI eq '' ){
Sebastien Moretti's avatar
Sebastien Moretti committed
351
352
                print "\tNo protein (prot GI) link found for $equivalent_blast_hits[$qq] in $fasta_header\n\n";
                @failureStatus = ('PUI_unavailable', $equivalent_blast_hits[$qq]);
353
                next HIT_LINK;
354
            }
Sebastien Moretti's avatar
Sebastien Moretti committed
355

Sebastien Moretti's avatar
Sebastien Moretti committed
356
            # PUID -> nt GIs & GeneID
Sebastien Moretti's avatar
Sebastien Moretti committed
357
            my ($ntGIs, $geneID) = protGI2NTGIs($protGI, $equivalent_blast_hits[$qq]);
358
            if ( $ntGIs eq '' && $geneID eq '' ){
Sebastien Moretti's avatar
Sebastien Moretti committed
359
360
                print "\tNo nucleotide (nt GIs) link found for $equivalent_blast_hits[$qq] in $fasta_header\n\n";
                @failureStatus = ('No_nt_link', $equivalent_blast_hits[$qq]);
361
                next HIT_LINK;
362
363
364
365
            }
            print " linked with [$ntGIs $geneID]\n";


Sebastien Moretti's avatar
Sebastien Moretti committed
366
            # Get transcript and contig seq
Sebastien Moretti's avatar
Sebastien Moretti committed
367
            if ( $ntGIs ne '' ){
Sebastien Moretti's avatar
Sebastien Moretti committed
368
                @nt_GIs = split(/,/, $ntGIs);
Sebastien Moretti's avatar
Sebastien Moretti committed
369
370
                downloadSeqFromGIs($cache, '', '', @nt_GIs);
            }
371

Sebastien Moretti's avatar
Sebastien Moretti committed
372
            # GeneID -> Chr
373
374
375
376
377
378
379
            if ( $geneID ne '' ){
                my @geneIDs = split(/,/, $geneID);
                while(<@geneIDs>){
                    my $geneID = $_;
                    my ($chr, $amont, $aval) = geneID2Chr($geneID, $equivalent_blast_hits[$qq]);
                    print "\n\tNo gene locus found for $equivalent_blast_hits[$qq] with $geneID in $fasta_header\n\n" if ($chr eq '' and $geneID ne '');

Sebastien Moretti's avatar
Sebastien Moretti committed
380
                    # Get Chr seq
381
382
                    if ( $chr ne '' ){
                        download_seq($cache, $amont, $aval, $chr);
Sebastien Moretti's avatar
Sebastien Moretti committed
383
                        $intronStep = 1;
384

385
                        unshift(@nt_GIs, "$chr--$amont-$aval");
386
                    }
387
388
389
390
391
                }
            }
        }


Sebastien Moretti's avatar
Sebastien Moretti committed
392
393
        # Align Nt seq with our protein query seq
        my ($POS, $BOJ) = runExonerate($cache, $date, $r, $original_seq[$r], $fasta_header, @nt_GIs);
394
        next HIT_LINK if ( !defined $POS );
395

Sebastien Moretti's avatar
Sebastien Moretti committed
396
        my ($POSresult, $bestNucleotide) = prepareResults4CDS($POS, $equivalent_blast_hits[$qq], $original_seq[$r], $fasta_header);
397
398
        $resultPOS .= $POSresult if ( $bestNucleotide ne '' && $POSresult ne '' && $resultPOS !~ /_G_$bestNucleotide/ );

399
400
        @intronStatus = ($POS->{0}, $equivalent_blast_hits[$qq], '')              if ( $intronStep==1 && exists($POS->{0}) && $POS->{0} =~ /\:/ );
        @intronStatus = ($POS->{0}, $equivalent_blast_hits[$qq], $bestNucleotide) if ( $intronStep==0 && $bestNucleotide =~ /^NA[CTWZGS]_/ );
401
402
403

        my ($intronLess, $nameLess) = ('', '');
        $intronLess = $POS->{0}       if ( %$BOJ eq 0 && exists($POS->{0}) && $POS->{0} ne '' );
404
        $nameLess   = $bestNucleotide if ( %$BOJ eq 0 && $intronStep==0 && $bestNucleotide =~ /^NA[CTWZGS]_/ );
Sebastien Moretti's avatar
Sebastien Moretti committed
405
        my ($BOJresult, $bestGenomic) = prepareResults4BOJ($BOJ, $equivalent_blast_hits[$qq], $original_seq[$r], $fasta_header, $intronLess, $nameLess);
406
        $resultBOJ .= $BOJresult      if ( $bestGenomic ne '' && $BOJresult ne '' && $resultBOJ !~ /_G_$bestGenomic[ \-]/ );
Sebastien Moretti's avatar
Sebastien Moretti committed
407
408
409
    }


410
    if ( $resultPOS eq '' ){
411
        if ( $failureStatus[0] ne '' ){
412
            buildFailureOutputFiles($r, $failureStatus[1],         $failureStatus[0]);
413
414
        }
        else {
415
            buildFailureOutputFiles($r, $equivalent_blast_hits[0], 'Alignment_failure');
416
        }
417
        next EACH_SEQ;
Sebastien Moretti's avatar
Sebastien Moretti committed
418
    }
419
420
    elsif ( $resultBOJ eq '' ){
        if ( $failureStatus[0] eq 'PUI_unavailable' || $failureStatus[0] eq 'No_nt_link' ){
421
            buildFailureBOJOutputFile($failureStatus[1], $failureStatus[0], $fasta_header, $original_seq[$r]);
422
423
        }
        elsif ( $failureStatus[0] eq '' && $intronStatus[0] ne '' ){
424
            buildIntronlessBOJOutputFile($intronStatus[1], $intronStatus[0], $intronStatus[2], $fasta_header, $original_seq[$r]);
425
426
        }
        else{
Sebastien Moretti's avatar
Sebastien Moretti committed
427
            buildFailureBOJOutputFile($equivalent_blast_hits[0], 'Unavailable', $fasta_header, $original_seq[$r]);
428
        }
429
        createCDSOutputFile($resultPOS, $original_seq[$r], $fasta_header);
Sebastien Moretti's avatar
Sebastien Moretti committed
430
431
    }
    else {
432
        createCDSOutputFile($resultPOS, $original_seq[$r], $fasta_header);
433
        createBOJOutputFile($resultBOJ);
Sebastien Moretti's avatar
Sebastien Moretti committed
434
435
436
437
438
439
440
441
    }


    undef $original_seq[$r];
    undef $original_names[$r];
}


442
checkAndCleanStderrFiles("$cache/$date.ExonerateError")  if ( -e "$cache/$date.ExonerateError" );
Sebastien Moretti's avatar
Sebastien Moretti committed
443
checkOutputFiles();
444
unlink("$msa")  if ( $converted==1 );
445
unlink 'blast_result.txt';
Sebastien Moretti's avatar
Sebastien Moretti committed
446
my @tmpFiles = glob($cache."/$date*");
Sebastien Moretti's avatar
Sebastien Moretti committed
447
if ( exists($tmpFiles[0]) ){
Sebastien Moretti's avatar
Sebastien Moretti committed
448
    mkdir("$cache/$date.TMP");
449
    move("$_", "$cache/$date.TMP")  while(<@tmpFiles>);
Sebastien Moretti's avatar
Sebastien Moretti committed
450
451
452
453
454
455
456
}


#Ouput Checker


#STDOUT Log output for our server
Sebastien Moretti's avatar
Sebastien Moretti committed
457
if ( -s "$originalMSA.cds" || -s "$originalMSA.out" ){
Sebastien Moretti's avatar
Sebastien Moretti committed
458
    print "\n\nTERMINATION STATUS: SUCCESS\n";
459
    print 'OUTPUT RESULTS', "\n";
460
    print "    #### File Type=        MSA Format= fasta_CDS Name= $originalMSA.cds\n"
461
        if ( -s "$originalMSA.cds" );
462
    print "    #### File Type=        MSA Format= fasta_CDS+Query Name= $originalMSA.cdsP\n"
463
        if ( -s "$originalMSA.cdsP" );
464
    print "    #### File Type=        MSA Format= colored_fasta_CDS Name= $originalMSA.cdsP.html\n"
465
        if ( Views::Html("$originalMSA.cdsP") && -s "$originalMSA.cdsP.html" );
466
    print "    #### File Type=        MSA Format= Rejected_seq Name= $originalMSA.out\n"
467
        if ( -s "$originalMSA.out" );
468
    print "    #### File Type=        MSA Format= fasta_Exon-Boundaries Name= $originalMSA.boj\n"
469
        if ( -s "$originalMSA.boj" && -s "$originalMSA.cds" && $hideBOJ==0 );
Sebastien Moretti's avatar
Sebastien Moretti committed
470
471

    unlink "$cache/web_tempo.result";
Sebastien Moretti's avatar
Sebastien Moretti committed
472
473
}
else{
Sebastien Moretti's avatar
Sebastien Moretti committed
474
    failure();
475
    exit(1);
Sebastien Moretti's avatar
Sebastien Moretti committed
476
477
}

478
exit(0);
Sebastien Moretti's avatar
Sebastien Moretti committed
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510


=head1 NAME

ProtoGene.pl - Converts a peptidic alignment to a nucleotidic alignment through database search

=head1 SYNOPSIS

B<ProtoGene.pl> --msa=path_of_the_fasta_msa_file [options]

with these options available:

=over 4

=item I<--orgm>=targeted species in the database search

=item All_organisms [default], Bacteria, Viruses, Vertebrata, Eukaryota, Mammalia, Primates, Homo_sapiens, Gallus_gallus, Bos_taurus, Escherichia_coli, Arabidopsis_thaliana, Mus_musculus, Drosophila_Melanogaster, ...

=item I<--db>=targeted database in the database search

=item nr, pdb, swissprot, refseq_protein [default]

=item I<--local> to execute a local BLAST query with  --db=path_for_a_local_db_blast_formated

=item I<--template> to provide your own nucleotidic sequences following the cds file format

=item I<--revtrans> to reverse-translate sequences with no blast hit, in IUB (IUPAC) depiction code

=item They are removed from the alignement by default

=item I<--pep> to add the original peptide query beneath the back-translated sequence

511
512
513
514
515
516
517
=item I<--cachedir>=your_own_path_directory

=item cachedir sets cache directory

=item I<--cacheclean>=none, update, use, old, empty

=item cacheclean manages cache behavior
Sebastien Moretti's avatar
Sebastien Moretti committed
518
519
520
521
522
523
524
525
526

=item none: no cache usage, none temporary files are stored

=item update: use cache but update old files [default]

=item use: force use cache, whatever the age of files

=item old: remove the old files in the cache directory

527
=item empty: empty completely the cache directory
Sebastien Moretti's avatar
Sebastien Moretti committed
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548

=item I<--version> to print version information

=item I<--help> to print this help message

=back

=head1 DESCRIPTION

PROTOGENE : Bona-Fide Back Translation of Protein Multiple Sequence Alignments
It converts a peptidic alignment to a nucleotidic alignment through database search.
Blast queries allow to get nucleotidic sequences from where peptidic sequences come from.
Exonerate aligns nucleotidic and peptidic sequences together.
PROTOGENE re-builds the original alignment with nucleotidic information it has gotten.

=head1 REQUIREMENT

=over 4

=item B<Perl 5.6 or better> is required !

549
=item Standard Perl modules B<lib>, B<strict>,  B<warnings>,  B<diagnostics> are required
Sebastien Moretti's avatar
Sebastien Moretti committed
550

551
=item as well as some other current ones : B<Getopt::Long>,  B<File::Which>,  B<File::Copy>, B<Mail::Send>, B<Time::localtime>, B<LWP::Simple>
Sebastien Moretti's avatar
Sebastien Moretti committed
552
553
554

=item -

555
=item I<exonerate> version 2 from http://www.ebi.ac.uk/about/vertebrate-genomics/software/exonerate
Sebastien Moretti's avatar
Sebastien Moretti committed
556

557
=item I<blast> from ftp://ftp.ncbi.nih.gov/blast/executables/
Sebastien Moretti's avatar
Sebastien Moretti committed
558
559
560
561
562
563
564

=back

=head1 VERSION

=over 8

565
=item version 4.2.2
Sebastien Moretti's avatar
Sebastien Moretti committed
566

567
=item on Sep 09th, 2016
Sebastien Moretti's avatar
Sebastien Moretti committed
568
569
570
571
572
573
574
575
576

=back

=head1 AUTHORS

=over 8

=item Sebastien MORETTI

Sebastien Moretti's avatar
Sebastien Moretti committed
577
=item moretti.sebastien [AT] gmail.com
Sebastien Moretti's avatar
Sebastien Moretti committed
578

579
=item Vital-IT computing center
Sebastien Moretti's avatar
Sebastien Moretti committed
580

581
=item Swiss Institute of Bioinformatics
Sebastien Moretti's avatar
Sebastien Moretti committed
582

583
=item Lausanne, Switzerland
Sebastien Moretti's avatar
Sebastien Moretti committed
584

585
=item http://www.vital-it.ch/
Sebastien Moretti's avatar
Sebastien Moretti committed
586
587
588
589
590
591
592
593

=back

=cut


######################  Management  ######################

594
# Failure declaration
Sebastien Moretti's avatar
Sebastien Moretti committed
595
sub failure {
596
597
598
    print {*STDERR} "\n\nFATAL\n\nTERMINATION STATUS: FAILURE\n\n";
    return;
}
Sebastien Moretti's avatar
Sebastien Moretti committed
599

Sebastien Moretti's avatar
Sebastien Moretti committed
600
# Check external programs presence in the PATH
601
sub checkExternalSoftware {
Sebastien Moretti's avatar
Sebastien Moretti committed
602
    for my $exe ( @_ ){
603
        if ( ! which($exe) && !-x $exe ){
Sebastien Moretti's avatar
Sebastien Moretti committed
604
605
606
607
            failure();
            print {*STDERR} "\t'$exe' program is not reachable\n\tIt could not be in your PATH or not installed\n\n";
            exit(1);
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
608
    }
Sebastien Moretti's avatar
Sebastien Moretti committed
609
    return;
Sebastien Moretti's avatar
Sebastien Moretti committed
610
611
}

612
613
# Check cache directory accessibility: Test and change rights, if required, of the cache directory
sub checkCacheAccessibility {
614
    my ($cacheDir) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
615

616
    if ( -d "$cacheDir/" && -w "$cacheDir/" && -x "$cacheDir/" && -r "$cacheDir/" ){
617
        # Good
Sebastien Moretti's avatar
Sebastien Moretti committed
618
    }
619
620
621
    elsif ( -d "$cacheDir/" ){
        if ( -o "$cacheDir/" ){
            chmod 0754, "$cacheDir/";
622
        }
623
        else{
624
625
626
            failure();
            print {*STDERR} "\tPermissions do NOT seem to be valid for the '$cacheDir' cache directory\n\n";
            exit(2);
627
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
628
629
    }
    else{
630
        mkdir "$cacheDir/"; # FIXME: mkdir is not recursive !
631
        if ( -o "$cacheDir/" ){
632
            chmod 0754, "$cacheDir/"; #0754 is in octal, so 754 or '0754' are not good !
633
634
        }
        else{
635
636
637
            failure();
            print {*STDERR} "\tPermissions do NOT seem to be valid for the '$cacheDir' cache directory\n\n";
            exit(2);
638
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
639
    }
640

641
    return;
Sebastien Moretti's avatar
Sebastien Moretti committed
642
643
}

644
# Cache Behavior Management
Sebastien Moretti's avatar
Sebastien Moretti committed
645
sub cacheManagement{
646
    my ($cache, $cacheDir, $inScript) = @_;
647

648
    if ( $cache eq 'empty' ){
649
        unlink <$cacheDir/*.fas>;
Sebastien Moretti's avatar
Sebastien Moretti committed
650
        unlink <$cacheDir/*.ExonerateError>;
651
652
        unlink("$cacheDir/webblast.log", "$cacheDir/web_tempo.result", "$cacheDir/debug.tempo");
        print "\n\tcache directory [$cacheDir] is empty\n\n";
653
        exit 0;
Sebastien Moretti's avatar
Sebastien Moretti committed
654
    }
655
    elsif ( $cache eq 'old' ){
656
657
658
659
660
        unlink    grep { -M $_ > $cacheStorageTime }    glob($cacheDir.'/*.fas'); # Grep list only with $cacheStorageTime days old files

        if ( $inScript==0 ){
            print "\n\tcache directory [$cacheDir] is empty of its old files\n\n";
            exit 0;
661
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
662
    }
663
    elsif ( $cache eq 'update' || $cache eq 'use' ){
664
665
        # DEFAULT: Limit cached files to less than $cacheStorageTime days old ones for 'update'
        $cache = $cacheDir;
Sebastien Moretti's avatar
Sebastien Moretti committed
666
    }
667
    elsif ( $cache eq 'none' ){
668
        # Do not use cache
Sebastien Moretti's avatar
Sebastien Moretti committed
669
    }
670
671
672
673
    elsif ( -d $cache ){
        die "\n\tProtoGene cannot access to your directory\n\n"  if ( ! -x $cache );
        die "\n\tProtoGene cannot write into your directory\n\n" if ( ! -w $cache );
        die "\n\tProtoGene cannot read into your directory\n\n"  if ( ! -r $cache );
Sebastien Moretti's avatar
Sebastien Moretti committed
674
675
    }
    else {
676
677
678
        failure();
        print {*STDERR} "\n\tWrong cache argument, or your cache folder '$cacheDir' doesn't seem to exist\n\n";
        exit(3);
Sebastien Moretti's avatar
Sebastien Moretti committed
679
680
681
682
683
684
685
686
687
688
    }

    return($cache);
}

##########################################################


######################    Fasta    ######################

Sebastien Moretti's avatar
Sebastien Moretti committed
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
# Check input aln format
sub isFastaFile{
    my ($msa, $converted) = @_;

    if ( -s "$msa" ){ # Check is the file exists and is not empty
        # Is FASTA format ?
        my $firstLine = `head $msa`;
        return($msa, 0) if ( $firstLine =~ /^>/ );

        # Convert no fasta format to fasta with readseq
        my $reformat = which('readseq') || '';
        if ( $reformat ne '' ){
            system("$reformat -a -f8 $msa > $msa.007")==0 or do {failure(); print {*STDERR} "\n\tEnable to convert your file to fasta format\n\n";};
            $msa      .= '.007';
            $converted = 1;
704
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
705
706
    }
    else {
Sebastien Moretti's avatar
Sebastien Moretti committed
707
        failure();
708
709
        print {*STDERR} "\tThe file \'$msa\' doesn't not seem to be reachable\n\n";
        exit(1);
Sebastien Moretti's avatar
Sebastien Moretti committed
710
711
    }

Sebastien Moretti's avatar
Sebastien Moretti committed
712
    return($msa, $converted);
Sebastien Moretti's avatar
Sebastien Moretti committed
713
714
715
716
}

#Prepare Fasta Header of the query name to incorporate our annotations
sub fastaHeaders4ProtoGene{
717
    my ($QueryName) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
718

719
    $QueryName =~ s{  +}{ }g;
720
    my ($acc, $desc) = ($QueryName, $QueryName);
721
722
    $acc  =~ s{^ *([^ ]*) .*$}{$1};
    $acc  =~ s{\s+$}{}g;
723
724
    $acc  =  $acc.'_G_@@';

725
726
727
    $desc =~ s{^ *[^ ]* *(.*)$}{$1};
    $desc =~ s{^[\| \.]+}{};
    $desc =~ s{[\|\. ]+$}{};
728
729

    return($acc.$desc);
Sebastien Moretti's avatar
Sebastien Moretti committed
730
731
732
733
}

#Prepare Fasta Header to put it on the fasta header of the result outputs
sub getFastaHeaderAnnot{
734
    my ($bestTarget) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
735
736

    #Get annotation from nucleotide file header
Sebastien Moretti's avatar
Sebastien Moretti committed
737
738
    open(my $BEST, '<', "$cache/$bestTarget.fas");
    my $annot = <$BEST> || '';
739
    close $BEST;
Sebastien Moretti's avatar
Sebastien Moretti committed
740

Sebastien Moretti's avatar
Sebastien Moretti committed
741
    chomp($annot);
742
743
744
745
    $annot =~ s{^ *[^ ]* }{};
    $annot =~ s{[\. ]*$}{};
    $annot =~ s{^ *}{};
    $annot =~ s{  +}{ }g;
Sebastien Moretti's avatar
Sebastien Moretti committed
746
747
748
749
750
751
752
753
754
755
756

    return($annot);
}

#########################################################


###################### Translation ######################

#Direct Reverse translation
sub reverse_trad{
757
    my ($aa) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
758

759
760
761
762
763
764
    my %reverse_code = ('A' => 'GCN', 'F' => 'TTy', 'K' => 'AAr', 'P' => 'CCN', 'T' => 'ACN',
                        'C' => 'TGy', 'G' => 'GGN', 'L' => 'yTN', 'Q' => 'CAr', 'V' => 'GTN',
                        'D' => 'GAy', 'H' => 'CAy', 'M' => 'ATG', 'R' => 'mGN', 'W' => 'TGG',
                        'E' => 'GAr', 'I' => 'ATh', 'N' => 'AAy', 'S' => 'wsN', 'Y' => 'TAy',
                        'X' => 'NNN', '-' => '---', '*' => 'Trr',
                       ); # '*' is for stop codon
Sebastien Moretti's avatar
Sebastien Moretti committed
765

766
    my $triplet = 'nnn';
767
    while( my ($x, $y) = each(%reverse_code) ){
768
769
770
771
        if ( uc($aa) eq $x ){
            $triplet = $y;
            last;
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
772
773
774
775
776
777
778
779
780
    }

    return($triplet);
}

#########################################################


####################### webblast #######################
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
sub display_blast_param {
    print "
             Program :                   BLASTp
             Evalue threshold :          $blast_param->{'evalue'}
             Matrix :                    $blast_param->{'matrix'}
             Filter :                    $blast_param->{'filter'}
             Equivalent best hits used : $blast_param->{'nbesthits'}

             Database:                   NCBI $blast_param->{'db1'}
             Blast_identity_threshold :  $blast_param->{'identity1'}
             Cover threshold :           $blast_param->{'cover1'}
          if no match
             Database:                   NCBI $blast_param->{'db2'}
             Blast_identity_threshold :   $blast_param->{'identity2'}
             Cover threshold :            $blast_param->{'cover2'}
\n";
    return;
}

sub run_BLAST{
    my ($infile, $whatBlast, $db, $blast_exe, $outfile) = @_;

    my $blast_at = $whatBlast==2  ? '-gigablast=yes'
                 : $whatBlast==1  ? "-blast_dir=$blast_exe"
                 :                  '';

    # BLAST run 1
    system("$webblast_exe -program=blastp -database=$db $blast_at -infile=$infile -matrix=$blast_param->{'matrix'} -evalue=$blast_param->{'evalue'} -method=geneid -filter=$blast_param->{'filter'} -organism=$species -identity=$blast_param->{'identity1'} -cover=$blast_param->{'cover1'} -hits=$blast_param->{'nbesthits'} -quiet=on -outfile=$outfile");


    if ( -z "$outfile" ){
    # BLAST run 2 if no hit with 100% identity and coverage: refseq100 -> nr95  OR  nr100 -> nr95
Sebastien Moretti's avatar
Sebastien Moretti committed
813
        print "run BLAST.. against $blast_param->{'db2'} because there was no acceptable hit against $blast_param->{'db1'}\n"
814
            if ( $whatBlast != 1 && $db ne 'nr' );
Sebastien Moretti's avatar
Sebastien Moretti committed
815
        print "run BLAST.. again, with decreased thresholds, because there was no acceptable hit against $db\n"
816
            if ( $whatBlast == 1 || $db eq 'nr' );
817
818

        system("$webblast_exe -program=blastp -database=$blast_param->{'db2'} $blast_at -infile=$infile -matrix=$blast_param->{'matrix'} -evalue=$blast_param->{'evalue'} -method=geneid -filter=$blast_param->{'filter'} -organism=$species -identity=$blast_param->{'identity2'} -cover=$blast_param->{'cover2'} -hits=$blast_param->{'nbesthits'} -quiet=on -outfile=$outfile");
Sebastien Moretti's avatar
Sebastien Moretti committed
819
820
    }

821
822
823
824
825
826
827
    move('webblast.log',     "$cache"); #Move temporary webblast.pl files
    move('web_tempo.result', "$cache");
    move('debug.tempo',      "$cache");

    unlink("$cache/webblast.log")     if ( -z "$cache/webblast.log" );
    unlink("$cache/web_tempo.result") if ( -z "$cache/web_tempo.result" );
    unlink("$cache/debug.tempo")      if ( -z "$cache/debug.tempo" );
Sebastien Moretti's avatar
Sebastien Moretti committed
828

829
    return;
Sebastien Moretti's avatar
Sebastien Moretti committed
830
831
832
833
834
835
836
}


########################################################


##################### NCBI requests #####################
837
838
839
sub fetch {
    my ($url) = @_;

840
    XML:
841
842
    for (my $tries=0; $tries <20; $tries++ ){
        my $content = get($url);
843
        print {*STDERR} "[$content]\n\n" if ($debug);
844
        if ( defined $content ){
845
            next XML  if ( $content =~ /<ERROR>/i && $content !~ /<ERROR>Can not find description/i );
846
847
            return $content;
        }
848
    }
Sebastien Moretti's avatar
Sebastien Moretti committed
849
    print {*STDERR} "\n\tERROR: Problem with NCBI eutils, please try again later\n";
850
    print {*STDOUT} "\n\tERROR: Problem with NCBI eutils, please try again later\n";
Sebastien Moretti's avatar
Sebastien Moretti committed
851
    exit(4);
852
}
853

854
855
856
sub fetch_fasta {
    my ($url, $outfile) = @_;

857
    return 1 if ( -s "$outfile" && $cache eq 'update' && -M "$outfile" <= $cacheStorageTime );
858

859
860
861
862
863
864
865
866
867
868
869
    GET_FASTA:
    for (my $tries=0; $tries <20; $tries++ ){
        my $HTTP_status = getstore($url, $outfile);
        if ( is_error($HTTP_status) ){
            next GET_FASTA;
        }
        else {
            open(my $FASTA, '<', "$outfile");
            my $counter = 0;
            my $lines   = 0;
            while(<$FASTA>){
870
#                $lines++              if ( $_ !~ /^>/ ); #Allow download of entries with only fasta header and no sequence associated
871
872
873
874
875
                $counter = $counter+2 if ( $counter==1 && $_ !~ /^\w/ && $lines==1 && $_ !~ /^>/ );
                $counter++            if ( $_ =~ /^>/ );
                $counter = $counter+2 if ( $_ !~ /^>/ && ($_ =~ /Error:/ || $_ =~ /[<>]/) );
            }
            close $FASTA;
876

877
878
879
            return 1 if ( $counter == 1 );
        }
    }
880
    print {*STDERR} "[[$url]]\n" if ( $debug );
881
    unlink("$outfile");
Sebastien Moretti's avatar
Sebastien Moretti committed
882
    print {*STDERR} "\n\tERROR: Problem with NCBI eutils, please try again later.\n";
883
    print {*STDOUT} "\n\tERROR: Problem with NCBI eutils, please try again later.\n";
Sebastien Moretti's avatar
Sebastien Moretti committed
884
    exit(5);
885
886
}

Sebastien Moretti's avatar
Sebastien Moretti committed
887
888
889

#Prot ACC -> PUID
sub blastPAcc2PGI{
890
    my ($blastHit) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
891

892
    my $protGI = '';
893
    #FIXME: should be ${blastHit}[pacc] but something is broken at NCBI
894
    my $content = fetch("https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=protein&term=$blastHit&retmode=xml&tool=ProtoGene&email=smoretti\@unil.ch");
895
896
    if ( $content =~ /<Id>(\d+)<\/Id>/ ){
        $protGI = $1;
Sebastien Moretti's avatar
Sebastien Moretti committed
897
898
899
900
901
902
903
    }

    return($protGI);
}

#PUID -> nt GIs
sub protGI2NTGIs{
904
    my ($protGI, $blastHit) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
905

906
907
    my $ntGIs  = '';
    my $geneID = '';
908
    my $content = fetch("https://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?dbfrom=protein&db=nuccore,gene&id=$protGI&retmode=xml&tool=ProtoGene&email=smoretti\@unil.ch");
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925

    my @xml  = split("\n", $content);
    my $flag = 0;
    for my $line (@xml){
        if (    $line =~ /<LinkName>protein_nuc[a-z]+<\/LinkName>/ ){ #for nuccleotide and nuccore
            $flag = 1;
        }
        elsif ( $line =~ /<LinkName>protein_gene<\/LinkName>/ ){
            $flag  = 2;
        }
        elsif ( $flag==1 && $line =~ /^.*<Id>(\d+)<\/Id>.*$/ ){
            my $match = $1;
            $ntGIs  .= "$match,$match,";
        }
        elsif ( $flag==2 && $line =~ /^.*<Id>(\d+)<\/Id>.*$/ ){
            my $match = $1;
            $geneID .= "$match,$match,";
926
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
927
928
    }

929
    #Remove redundancy if any
930
    chop $ntGIs; #Remove last ','
931
932
933
934
935
936
    my %hash_NT = split(',', $ntGIs);
    $ntGIs      = join(',',  keys(%hash_NT) );
    chop $geneID;
    my %hash_GE = split(',', $geneID);
    $geneID     = join(',',  keys(%hash_GE) );

937
    return($ntGIs, $geneID);
Sebastien Moretti's avatar
Sebastien Moretti committed
938
939
940
}

sub geneID2Chr{
941
    my ($geneID, $blastHit) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
942

943
944
    my $chr   = '';
    my ($amont, $aval) = ('', '');
945
946
#   my $content = fetch("https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=gene&id=$geneID&retmode=xml&tool=ProtoGene&email=smoretti\@unil.ch");
    my $content = fetch("https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=gene&id=$geneID&retmode=xml&tool=ProtoGene&email=smoretti\@unil.ch");
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965

    my @xml  = split("\n", $content);
    my $flag = 0;
    GENE_INFO:
    for my $line (@xml){
        if ( $flag==0    && $line =~ /<Item Name="GenomicInfoType" Type="Structure">/ ){
            $flag = 1;
        }
        elsif ( $flag==1 && $line =~ /<Item Name="ChrAccVer" Type="String">([^<]+?)\.?\d*<\/Item>/ ){
            $chr   = $1;
            $flag  = 3;
        }
        elsif ( $flag==3 && $line =~ /<Item Name="ChrStart" Type="Integer">(\d+)<\/Item>/ ){
            $amont = $1;
            $flag  = 4;
        }
        elsif ( $flag==4 && $line =~ /<Item Name="ChrStop" Type="Integer">(\d+)<\/Item>/ ){
            $aval  = $1;
            last GENE_INFO;
966
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
967
    }
968

969
970
    if ( $amont eq '' || $aval eq '' ){
        $amont = '';
971
        $aval  = '';
Sebastien Moretti's avatar
Sebastien Moretti committed
972
    }
973
974
    elsif ( $amont ne '' && $aval ne '' ){
        if ( $amont > $aval ){
975
976
            $amont = $amont + 5000;
            $aval  = $aval  - 5000;
977
978
        }
        else{
979
980
            $amont = $amont - 5000;
            $aval  = $aval  + 5000;
981
        }
Sebastien Moretti's avatar
Sebastien Moretti committed
982
    }
983
    $amont = 1 if ( $amont ne '' && $amont <= 0 );
984
    $aval  = 1 if ( $aval  ne '' && $aval  <= 0 );
Sebastien Moretti's avatar
Sebastien Moretti committed
985

986
    return($chr, $amont, $aval);
Sebastien Moretti's avatar
Sebastien Moretti committed
987
988
989
}

sub downloadSeqFromGIs{
990
    my ($cache, $amont, $aval, @acc) = @_;
Sebastien Moretti's avatar
Sebastien Moretti committed
991

992
    GET_SEQ: