$sequence = preg_replace("/\\W|\\d/", "", $sequence);
// maximum length of amplicons
$maxlength = $_POST["length"];
// SET PATTERNS FROM PRIMERS
// Change N to point in primers
$pattern1 = str_replace("N", ".", $primer1);
$pattern2 = str_replace("N", ".", $primer2);
// If one missmatch is allowed, create new pattern
// example: pattern="ACGT"; to allow one missmatch pattern=".CGT|A.GT|AC.T|ACG."
if ($_POST["allowmissmatch"] == 1) {
$pattern1 = includeN($primer1);
$pattern2 = includeN($primer2);
}
// SET PATTERN
$start_pattern = "{$pattern1}|{$pattern2}";
$end_pattern = RevComp($start_pattern);
// CALL Amplify FUNCTION
$results_array = Amplify($start_pattern, $end_pattern, $sequence, $maxlength);
// PRINT RESULTS
print "<pre>Primer 1: {$primer1}\\n";
print "Primer 2: {$primer2}\\n\\n";
if (sizeof($results_array) > 0) {
print "List of amplicons: position in sequence, length and sequence\\n\\n";
foreach ($results_array as $key => $val) {
print "{$key}\t{$val}\t" . substr($sequence, $key, $val) . "\\n";
}
} else {
print "No amplification\\n\\n";
}
// ##############################################################
// FUNCTIONS
function compute_zscores_for_tetranucleotides($theseq)
{
// as described by Teeling et al. BMC Bioinformatics 2004, 5:163.
$theseq .= " " . RevComp($theseq);
$i = 0;
$len = strlen($theseq) - 2 + 1;
while ($i < $len) {
$seq = substr($theseq, $i, 2);
$oligos2[$seq]++;
$i++;
}
$i = 0;
$len = strlen($theseq) - 3 + 1;
while ($i < $len) {
$seq = substr($theseq, $i, 3);
$oligos3[$seq]++;
$i++;
}
$i = 0;
$len = strlen($theseq) - 4 + 1;
while ($i < $len) {
$seq = substr($theseq, $i, 4);
$oligos4[$seq]++;
$i++;
}
$base_a = array("A", "C", "G", "T");
$base_b = array("A", "C", "G", "T");
$base_c = array("A", "C", "G", "T");
$base_d = array("A", "C", "G", "T");
$base_e = array("A", "C", "G", "T");
$base_f = array("A", "C", "G", "T");
// COMPUTE Z-SCORES FOR TETRANUCLEOTIDES
$i = 0;
foreach ($base_a as $key_a => $val_a) {
foreach ($base_b as $key_b => $val_b) {
foreach ($base_c as $key_c => $val_c) {
foreach ($base_d as $key_d => $val_d) {
$exp[$val_a . $val_b . $val_c . $val_d] = $oligos3[$val_a . $val_b . $val_c] * $oligos3[$val_b . $val_c . $val_d] / $oligos2[$val_b . $val_c];
$var[$val_a . $val_b . $val_c . $val_d] = $exp[$val_a . $val_b . $val_c . $val_d] * (($oligos2[$val_b . $val_c] - $oligos3[$val_a . $val_b . $val_c]) * ($oligos2[$val_b . $val_c] - $oligos3[$val_b . $val_c . $val_d]) / pow($oligos2[$val_b . $val_c], 2));
$zscore[$i] = ($oligos4[$val_a . $val_b . $val_c . $val_d] - $exp[$val_a . $val_b . $val_c . $val_d]) / sqrt($var[$val_a . $val_b . $val_c . $val_d]);
$i++;
}
}
}
}
return $zscore;
}
function FCGR_compute($input_min, $input_max)
{
print "<p align=right><a href=" . $_SERVER["PHP_SELF"] . ">Home</a></p>\n";
print "Computing...(time depends on sequence length and power of the server)<center><hr>";
flush();
// GET DATA
if ($_POST["seq_name"]) {
$seq_name = $_POST["seq_name"];
} else {
$seq_name = "No name";
}
$seq = strtoupper($_POST["seq"]);
$seq = preg_replace("/\\W|\\d/", "", $seq);
$seq_len = strlen($seq);
// limits for length of sequence
if ($seq_len > $input_max) {
die("<p>Sequence is longer than {$input_max} bp.<p>At this moment we can not provide this service to such a long sequences.");
}
if ($seq_len < $input_min) {
die("<p>Minumum sequence length: {$input_min} bp");
}
$oligo_len = $_POST["len"];
// If double strand is requested to be computed...
if ($_POST["s"] == 2) {
$seq .= " " . RevComp($seq);
}
// compute nucleotide frequencies
$A = substr_count($seq, "A");
$C = substr_count($seq, "C");
$G = substr_count($seq, "G");
$T = substr_count($seq, "T");
// COMPUTE OLIGONUCLEOTIDE FREQUENCIES
// frequencies are saved to an array named $oligos
$oligos = find_oligos($seq, $oligo_len);
// CREATE CHAOS GAME REPRESENTATION OF FREQUENCIES IMAGE
// check the function for more info on parameters
// $data contains a string with the data to be used to create the image map
$for_map = create_FCGR_image($oligos, $seq_name, $A, $C, $G, $T, $seq_len, $_POST["s"], $oligo_len);
// PRINT THE IMAGE, WHICH WILL BE A IMAGE MAP WHEN REQUESTED
// to avoid submission of a huge amount of data throught the net
if ($_POST["map"] == 1) {
// image map is requested
print "<br><MAP NAME=Kaixo>\n{$for_map}\n</MAP>\n<img USEMAP=\\#Kaixo src=FCGR.png?" . date("U") . " width=552 hight=700 border=0>\n";
} else {
print "<br><img src=FCGR.png?" . date("U") . " width=552 hight=700 border=0>";
}
// PRINT TEXTAREA WITH OLIGONUCLEOTIDE FREQUENCIES WHEN REQUESTED
if ($_POST["freq"] == 1) {
// oligonucleotide frequencies are requested
print "<p><p>Raw data used to generate images above: <BR><textarea cols=80 rows=10>Sequence\tOccurences\n";
foreach ($oligos as $key => $val) {
print "\n{$key}\t{$val}";
}
print "</textarea>";
}
}
$sequence = preg_replace("/\\W|\\d/", "", $sequence);
// removed
// when length of query sequence is 0 => error
if (strlen($sequence) == 0) {
die("Error: query sequence not provided. Plase go back andtry again.");
}
if (strlen($sequence) > 1000000) {
die("Error: sequence is too long. Download the script from biophp.org and used it localy.");
}
// when length of query sequence is bellow 4^oligo_len => error (to avoid a lot of 0 frequencies);
if (strlen($sequence) < pow(4, $oligo_len)) {
die("Error: query sequence must be at least 4^(length of oligo) to proceed.");
}
// when frequencies at both strands are requested, place sequence and reverse complement of sequence in one line
if ($strands == 2) {
$sequence .= " " . RevComp($sequence);
}
// compute request and save data in an array
$result = find_oligos($sequence, $oligo_len);
// print the form
print_form($sequence);
//print out results
print "<p>Frequencie of oligos with length {$oligo_len}<br><textarea cols=60 rows=50>";
foreach ($result as $oligo => $frequency) {
print "{$oligo}\t{$frequency}\n";
}
print "\n</textarea>\n";
}
// ######################################################################################################
// ##################################### FUNCTIONS ###################################
// ######################################################################################################
