Nextflow ProteinFunction: support for SQLite db

[nakib103]

ENSVAR-6567

Extend ProteinFunction nextflow pipeline to be able to write to a SQLite db. It introduces the following params -

params.offline - if set, no ensembl database connection would be made (and not storing data in MySQL db)
params.sqlite - if set, a SQLite db would be created with the results. (if params.offline is set params.sqlite is automatically set too, otherwise there would be no output)
params.sqlite_dir - directory location where sqlite db should be stored. By default it is the param.outdir. The db name is set to ${params.species}_PolyPhen_SIFT.db
params.sqlite_db - give full path of the SQLite db with name. Alternative way to give db name.

Test:

Nextflow command -

nextflow run $ENSEMBL_ROOT_DIR/ensembl-variation/nextflow/ProteinFunction \
-profile $profile \
--sift_run_type FULL \
--outdir $PWD/temp \
--species felis_catus \
--offline 1 \
--gtf <gtf file>
--fasta <fasta file>

Test the generated SQLite db ($PWD/temp/felis_catus_PolyPhen_SIFT.db) against what we have in Ensembl database.

Test script

use DBI;
use Data::Dumper;
use Bio::EnsEMBL::Variation::ProteinFunctionPredictionMatrix;
use Bio::EnsEMBL::Registry;

$sqlite_db = "/some/path/felis_catus_PolyPhen_SIFT.db";

my $dbh = DBI->connect("dbi:SQLite:dbname=$sqlite_db","","");

my $registry = 'Bio::EnsEMBL::Registry';
$registry->load_registry_from_db(
    -host => 'HOST',
    -user => 'USER',
    -port => PORT,
    -pass => '',
    -db_version   => 110
);
$registry->add_alias('felis_catus', 'cat');

my $vdb = $registry->get_DBAdaptor('cat', 'variation');
my $pfpma = $vdb->get_ProteinFunctionPredictionMatrixAdaptor();

my $ITER = 5;
my $sth = $dbh->prepare("SELECT * from predictions LIMIT $ITER;");
$sth->execute();

foreach my $iter (1..$ITER) { 
    my $row = $sth->fetch();
    my $pfpm = Bio::EnsEMBL::Variation::ProteinFunctionPredictionMatrix->new(
        -analysis     => 'sift',
        -matrix       => $row->[2],
    );

    my $pfpm_test = $pfpma->fetch_sift_predictions_by_translation_md5($row->[0]);


    my $AAs = ['A', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'K', 'L', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'V', 'W', 'Y'];
    my $times = 0;
    my $POS = scalar keys $pfpm->deserialize;
    foreach my $aa (@{ $AAs }) {
        foreach my $pos (1..$POS) {
            my ($pred, $score) = $pfpm->prediction_from_matrix($pos, $aa);
            my ($pred_test, $score_test) = $pfpm_test->prediction_from_matrix($pos, $aa);

            if ($pred ne $pred_test && $score ne $score_test) {
                my $pred_match = $pred eq $pred_test ? "pred_match" : "pred_nomatch";
                print "$pos - $aa; $pred_match ; ", $score - $score_test, "\n";
            }
        }
    }
}

[likhitha-surapaneni]

Hi @nakib103 , we may need to define sift_run when params.sift_run_type != "NONE"

[nakib103]

Thanks! it has been fixed.