Filtering BLAST results
2
0
Entering edit mode
5.8 years ago
flogin ▴ 280

I have a csv file like this:

subject,query

VBAnDirWR_KB672490,INVdsDNA_NC_029032

VBAnDirWR_KB672490,VIRdsDNA_NC_029032

VBAnDirWR_KB672490,VIRdsDNA_NC_029032

VBAnDirWR_KB672490,VIRdsDNA_NC_029032

VBAnDirWR_KB672490,VIRdsDNA_NC_029032

VBAnDirWR_KB672490,BACdsDNA_NC_029032

VBAnDirWR_KB672491,VIRdsDNA_NC_029032

VBAnDirWR_KB672491,VIRdsDNA_NC_029032

VBAnDirWR_KB672491,BACdsDNA_NC_029032

VBAnDirWR_KB672491,BACdsDNA_NC_029032

VBAnDirWR_KB672491,INVdsDNA_NC_029032

VBAnDirWR_KB672492,BACdsDNA_NC_029032

VBAnDirWR_KB672492,VIRdsDNA_NC_029032

VBAnDirWR_KB672492,VIRdsDNA_NC_029032

VBAnDirWR_KB672492,VIRdsDNA_NC_029032

VBAnDirWR_KB672492,VIRdsDNA_NC_029032

As we can see, are 5 queries per subject (results of -num_alignments 5, in BLAST), I need recovery only subjects IDs that have the 'VIR' tag in 3 or more queries, so I wrote (or I tried wrote xD), a python script to do this:

import pandas as pd
list_subjects = []
vir = 0
count = 0

# loading archive...
blast = pd.read_csv('blast_teste.csv')
# converting in a data frame...
blast_df = pd.DataFrame(blast)
# separing columns in subject and query
subject = blast_df.iloc[:,0]
query = blast_df.iloc[:,1]
# reading subject and query at the same time...
for i,c in zip(subject,query): 
    if count < 5: # to count viral tag each five elements.
        if vir < 3 and 'VIR' in c: # if tag in query
            vir += 1 
            count += 1
        elif vir >= 3: # if each 5 queries, 3 are viral, retrieve the subject ID
            list_subjects.append(i)
            count += 1
        else:
            count += 1
    elif count == 5: # when count reach in five, the count is reseted to start a new step of analysis
        count == 0
        vir == 0
print(list_subjects)

but this code returns: ['VBAnDirWR_KB672490'] and should be ['VBAnDirWR_KB672490','VBAnDirWR_KB672492'] (because those 2 subjects have more than 3 queries with 'VIR' tag.

can anybody help me?

alignment blast python • 1.4k views
ADD COMMENT
0
Entering edit mode

Thanks guys, both suggestions are working!

Best,

ADD REPLY
4
Entering edit mode
5.8 years ago
h.mon 35k

In Perl, using the same logic suggested by Carambakaracho .

#!/usr/bin/perl

use strict;
use warnings;

my %filter = ();
while (<>){
    my( $subject, $query ) = split ",";
    if( $query =~ /^VIR/ ){ $filter{$subject}++; }
}

foreach my $key ( keys %filter ){
    if( $filter{$key} > 2 ){ print "$key\n" };
}
ADD COMMENT
3
Entering edit mode
5.8 years ago
Carambakaracho ★ 3.3k

This looks much like some sort of constructed problem, does it? (Not to say "Homework") In any way, you're half way there - You need to reset the count variable, though it's quite useless. Simpler convert list_subjects to a dictionary, subject as key, 'VIR' count as value, increment each 'VIR' count per subject, return subjects with count >= 3

ADD COMMENT

Login before adding your answer.

Traffic: 1807 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6