Fails if all reads have common call at position

[keiranmraine]

PR and test example to follow, but here to allow head start on confirming I've understood this correctly @davidrajones

The following block has a problem:

https://github.com/cancerit/cgpCaVEManPostProcessing/blob/c520edd24a75a7d0c530e54129d52be43e2dfc0e/lib/Sanger/CGP/CavemanPostProcessor.pm#L308-L320

In the first block, where it tests for overlapping reads, it only records a result if both reads disagree:

https://github.com/cancerit/cgpCaVEManPostProcessing/blob/c520edd24a75a7d0c530e54129d52be43e2dfc0e/lib/Sanger/CGP/CavemanPostProcessor.pm#L309-L310

But it assigns the count only to the +ve strand:

https://github.com/cancerit/cgpCaVEManPostProcessing/blob/c520edd24a75a7d0c530e54129d52be43e2dfc0e/lib/Sanger/CGP/CavemanPostProcessor.pm#L311-L312

No counts are incremented if the reads agree which results in the process failing.

I think this should be as follows as the subsequent section handles the balancing of +/- strand:

foreach my $readnom( @$readname_arr){
    if(exists $hashed_reads->{$readnom}->{1} && exists $hashed_reads->{$readnom}->{-1}){
        $loc_counts->{1}->{qbase}++;
        $loc_counts->{-1}->{qbase}++;
    }elsif(exists $hashed_reads->{$readnom}->{1}){ # + strand populated only
        $loc_counts->{1}->{qbase}++;

    }else{ # - strand populated only
        $loc_counts->{-1}->{qbase}++;
    }
} # End of iteration through each readname for an initial count

Example data that triggered the problem:

$VAR1 = {
          'HS16_6996:1:1102:12569:60118#2' => {
                                                '-1' => {
                                                          'ln' => 75,
                                                          'rdPos' => 13,
                                                          'primaryalnscore' => 70,
                                                          'rdName' => 'HS16_6996:1:1102:12569:60118#2',
                                                          'softclipcount' => 0,
                                                          'qbase' => 'C',
                                                          'matchesindel' => '',
                                                          'xt' => undef,
                                                          'qual' => 60,
                                                          'qscore' => 9,
                                                          'str' => -1,
                                                          'start' => 146972148
                                                        },
                                                '1' => {
                                                         'str' => 1,
                                                         'start' => 146972110,
                                                         'qscore' => 42,
                                                         'qual' => 60,
                                                         'xt' => undef,
                                                         'qbase' => 'C',
                                                         'matchesindel' => '',
                                                         'softclipcount' => 0,
                                                         'rdName' => 'HS16_6996:1:1102:12569:60118#2',
                                                         'primaryalnscore' => 70,
                                                         'rdPos' => 51,
                                                         'ln' => 75
                                                       }
                                              },
          'HS12_6924:5:1206:17230:96528#2' => {
                                                '-1' => {
                                                          'rdName' => 'HS12_6924:5:1206:17230:96528#2',
                                                          'primaryalnscore' => 65,
                                                          'rdPos' => 16,
                                                          'ln' => 75,
                                                          'start' => 146972145,
                                                          'str' => -1,
                                                          'qual' => 60,
                                                          'xt' => undef,
                                                          'qscore' => 39,
                                                          'matchesindel' => '',
                                                          'qbase' => 'C',
                                                          'softclipcount' => 0
                                                        },
                                                '1' => {
                                                         'softclipcount' => 0,
                                                         'qbase' => 'C',
                                                         'matchesindel' => '',
                                                         'qual' => 40,
                                                         'xt' => undef,
                                                         'qscore' => 4,
                                                         'start' => 146972086,
                                                         'str' => 1,
                                                         'ln' => 75,
                                                         'rdPos' => 75,
                                                         'primaryalnscore' => 72,
                                                         'rdName' => 'HS12_6924:5:1206:17230:96528#2'
                                                       }
                                              }
        };

[keiranmraine]

Okay, so the read selection code needs to test the qscore when selecting the read to use. In both cases above the score is good in one read and poor in the other. Selecting the read with the highest qscore will fix the edge case.

[keiranmraine]

To be closed with #36

[keiranmraine]

Will be released once #37 changes are approved.