sorting confusion

[lilly07]

I have a tab separated text file which is as follows:

000017        chr4
000034        chr11
000035        chrY
000038        chr4
000040        chr4
000041        chr20
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrY
000051        chrX_random
000051        chrX_random
000051        chrX_random
023826        chr2
023827       chrY
023998        chr7
024100        chr6
024157        chr15
024245        chrY
024446        chrY
025091        chr2
025204        chrY
025431        chr19
024637        chr12
024834        chrY
024940        chrY
025747        chrY
026217        chr5
026398        chr18
026912        chr14
 

Basically I want to see whether in each record, left column eg( 026398) is having chrY on the right column and 026398 is a single record in the whole of left column.

For example, 026912 contains only chr14, hence a unique record where as 000051 consists of many hits on the right hand side. I tried using some unique commands but bit confusing.
Please let me know the logic

[nithinpes]

You can read the contents of the file, line by line and split each line across tab. You can check if the second element in the resulting array is the right-column value that you are looking for.
To determine uniqueness, you can make use of a hash with left-column value as key ($hash{$leftvalue}++). For any further assistance, you have to post the code that you tried.

[Ganon11]

Going a bit further with nithinpes' idea, you could probably have a hash whose keys are the left hand side numbers (i.e. 026912) and whose values are lists of results. For example:

%hash = (
   026912 => [chr14], # 026912 had 1 hit
   000051 => [chrY, chrY, chrY, chrX_random, chrX_random], # 000051 had 5 hits
   # etc...
);

To check uniqueness, just check the size of the array - if it is 1, that element is unique.

[lilly07]

Thanks, I could managed to add the value into a hash and it works fine till that. This is the first time, I am handling hash. Regarding the has manipulation, on how to check whether the array size of $value element is more than one ? I was able to find only the size of a hash table and couldn't find a way to manupulate among hash value or its size. Knidly let me know. Thanks.

#!/usr/bin/perl
my %hashcount = ();
 
while(<>) {
        chomp $_;
        s/\s+/\t/g;
        my(@v) = split(/\t/,$_);
        $hashcount{$v[0]} .= $v[1];
}
 
while( my ($k, $v) = each %hashcount ) {
        print "key: $k, value: $v\n";
    }

[lilly07]

I tried adding values into an array and then trying to find the array size. But it aways prints the array size as 1.

@value = $v;
              $value = @value;
        print "The array size: $value\n";
        if($value >1) {
        print "MULTIPLE HITS \n";
        }

Hence I think the way I add elements into hash itself is giving me problem. How do I over come this? Thanks.

[lilly07]

Hi I managed to solve the problem. I added a comma while adding values to the hash and using plit function to get an array of elements while retrieving as below.

#!/usr/bin/perl
my %hashcount = ();
 
 
while(<>) {
        chomp $_;
        s/\s+/\t/g;
        my(@v) = split(/\t/,$_);
        $hashcount{$v[0]} .= $v[1]."," ;
}
 
while( my ($k, $v) = each %hashcount ) {
        print "key: $k, value: $v\n";
        @value = split(/,/,$v);
              $value = @value;
        print "The array size: $value\n";
        if($value == 1){
 
                if( @value[0] eq "chrX") {
                print"UNIQUE HIT AT X\n";
                }else{
                print"UNIQUE HIT AT  NON X\n";
                }
 
 
        }else {
        print"MULTUPLE HITS \n";
}
 
}
 
 

Thanks. Now my objective is to find the elements in @value. I need to check whether all the elements in @value are ChrX or mix oc ChrY and others or does not have chrX at all. Is there any short cut method to achieve this? Thanks.

[KevinADC]

Easier to use a hash of arrays:

my %hashcount = ();
 
while(<>) {
        chomp;
        my @v = split(/\s+/);
        push @{$hashcount{$v[0]}}, $v[1];
}
 
while( my ($k, $v) = each %hashcount ) {
        print "key: $k, value: @{$v}\n";
}
 

Each hash key will have an array as its value instead of a string. You can check the size of the array associated with each hash key and/or loop/grep through the arrays to find whatever it is you need to find in them.

[KevinADC]

If you don't know about references yet then this code will look a little odd to you but they are not hard to learn.

my %hashcount = ();
 
 
while(<>) {
        chomp $_;
        my @v = split(/\s+/,$_);
        push @{$hashcount{$v[0]}}, $v[1];
}
 
while( my ($k, $v) = each %hashcount ) {
        print "key: $k, value: @{$v}\n";
        print "The array size: ", scalar @{$v}, "\n";
        if (@{$v} == 1){
                if( $v->[0] eq "chrX") {
                   print"UNIQUE HIT AT X\n";
                }else{
                   print"UNIQUE HIT AT  NON X\n";
                }
 
 
        }else {
        print"MULTUPLE HITS \n";
}
 
}
 

[lilly07]

Thanks Kevin. I learnt something new today.