change PlasMapper by cirdna EMBOSS

changed the description

Here are a number of blast parsers that I can consider:

from kirill kryukov
a complete solution is described on the bioperl site. It uses perl BlastResult objects

But calling blast with the appropriate output format option may be enough.

blastn -outfmt 6

This produces an output in tabular form:

mysequencethatilove	t3	100.000	41	0	0	122	162	1	41	1.43e-19	76.8
mysequencethatilove	t4	97.619	42	0	1	22	111	1	42	6.64e-18	71.3

with fields:

Query sequence (mysequencethatilove)
Subject name (t3)
Percent Identity (100.000)
Length of hit (41)
xxx(0)
Number of gaps (0)
Start on query (122)
End on query(162)
Start on subject (1)
End on subject (41)
E value (1.43e-19)
Score (76.8)

So, we need a routine that transforms this text in this format:

group

label
Block    122   162    3
t3
endlabel
label
Block    22   111    3
t4
endlabel

endgroup

Note that the last digit (3) indicates the color, we will implement it later.

I propose to address this issue with the following fonction :


stephane@berthollet:~$ nano -c blasttocirdna.php

<?php


/**
 *
 * This function allows you to import a blast outpout TXT file and export it into a new TXT file with cirdna input format
 *
 * @param string $file      The file you want to import the data from
 *
 * @return $outfile          The file you want to export the data to
 */


function transform_txt($file, $outfile)
{

//      echo "je suis dans la fonction "."\n";
        $newtxt = '';
        if(!$input = fopen($file, "r")){
                echo("Unable to open data file!");
                exit;
                }
        if(!$output = fopen($outfile, "w")){
                echo("Can't open file for writing!");
                exit;
                }
        fwrite($output, $newtxt);
        $txt_string = file_get_contents($file);
        $txt_string = trim($txt_string, "\n");   // trim() function removes whitespace
        $lines = explode("\n", $txt_string);

        $newtxt = "group" . "\n" . "\n"; fwrite($output, $newtxt);
//      $array = array();

         // For all the lines within the TXT
        foreach ($lines as $newline) {

                $line = explode("\t", $newline);
//            echo $line."\n";
//            print_r($line);

                echo  "Block". "\t" . $line["1"].  "\t" . $line["6"]. "\t" . $line["7"] . "\n";


                $newtxt =  "label" . "\n". "Block". "\t";  fwrite($output, $newtxt);
                $newtxt = $line["6"]. "\t" . $line["7"] . "\t" . "3" . "\n"; fwrite($output, $newtxt);
                $newtxt = $line["1"] . "\n"; fwrite($output, $newtxt);
                $newtxt =  "endlabel" . "\n" ; fwrite($output, $newtxt);
//              echo $newtxt . "\n";
/*
                if(fwrite($output, $newtxt) === FALSE) {
                        echo("Can't write to file!");
                        exit;
                 }
*/
        }
        $newtxt =   "\n". "endgroup"; fwrite($output, $newtxt);
//      fwrite($output, $newtxt. "\n". "endgroup");

        fclose($input);
        fclose($output);
}


$infile = 'blast/cequetuveux.txt'; // nom de mon fichier
$newdate = date('Y-m-d_H:i:s');
//echo $newdate;
$outfile =  'blast/inputcirdna_'.$newdate.'.txt';
echo $outfile . "\n";
transform_txt($infile,$outfile);   //calling function 


?>

to execute the script :

stephane@berthollet:~$ php blasttocirdna.php

mentioned in issue #3 (closed)

mentioned in issue #87 (closed)

mentioned in issue #75 (closed)

mentioned in commit f4cdb445

mentioned in commit 030a0620

mentioned in commit e0f89553

mentioned in commit 0d41609a

mentioned in commit 9bafe542

It works. We now need to:

optimize the size of blocks and text
deal with the fact that multiple users may write temporary files when browsing: create subdirs in storage/tmp/ with random long names so that each visitor writes there and only there. Then rm these dirs also.
check that $seq is dna
Handle block colors according to the type of features
Make a controller+route to update dna features blastdb (from a button in the browse view of dna_features, as we used to do it for plasmapper)

mentioned in commit 18040735

For random dirname, see this discussion

OK, tmpdir with random name is now created. We must remove it after usage: when, where to write this removal?

mentioned in commit 95ef0348

mentioned in commit 3a033232

CHANGED:

Instead, I decide to create the following tmpdir:

storage/app/public/tmp/plasmidmaps/id

where id is the id of the plasmid.

This way, we don't need to remove the tmpdirs. They are simply updated at every visit of the read view.

Regarding other points:

Optimize size of blocks and text. This is now better in svg format.
check that $seq is dna. I open new issue #105 for this.
block color according to feature type. I open a new issue #106 (closed) for this.
Make a controller+route to update dna features blastdb. I open new issue #107 (closed) for this.

And I close this issue.

closed

changed milestone to %first upgrade of gylab production site

mentioned in merge request !1 (merged)

change PlasMapper by cirdna EMBOSS

Designs

Child items ...

Activity