#!/usr/bin/perl

use 5.006;
use strict;
use warnings;
use File::Basename;

my $usage="\nperl $0 sample.gatk/vardict.hg19_multianno.txt
Input ANNOVAR table outputs and get Mutations statistic ";
die $usage unless @ARGV;

my $in_file=shift;#$sam.".gatk.hg19_multianno.txt";
my $sam=basename $in_file;
$sam=~s/\.hg19_multianno.txt//;
my $out_file=$sam.".stat.xls";
my (@samples,%data,@column,@sp_column,%indel);
open my $in_fh,'<',$in_file or die "Cannot open file:$!";

while(<$in_fh>){
    chomp;
    if(/^Chr\s+/ or $.==1){
        my @title=split /\t/;
        for my $i(0..$#title){
            if($title[$i] eq "Otherinfo13" or $title[$i] eq 'Otherinfo12'){
                    push @column,$i;
                }elsif($title[$i]=~/avsnp\d+$/ or $title[$i]=~/dbSNP/){                    
                    push @column,$i;
                }elsif($title[$i] eq 'gnomAD_genome_ALL'){                  
                    push @column,$i;
                }elsif($title[$i] eq 'Func.refGene'){                   
                    push @column,$i;
                }elsif($title[$i] eq 'ExonicFunc.refGene'){                 
                    push @column,$i;
                }elsif($title[$i] eq 'ALTS' or $title[$i] eq 'Alt'){
                    push @column,$i;
                }elsif($title[$i] eq 'Ref'){
                    push @column,$i;
                }
            }
        next;
    }
    my ($ref,$alt,$func,$efunc,$rs,$tg,$format,$sample)=(split /\t/)[@column]; ## must in order
    next if $sample =~/^\./;
    my @fmts=split /:/,$format;
    my $n_gt;
    for my $i (0..$#fmts){
        $n_gt=$i if $fmts[$i] eq 'GT';
    }
    my $gt=(split /:/,$sample)[$n_gt];
    next if $gt =~/^\./ or $gt eq '0/0';
    if($ref=~/[ATCG]/ and $alt=~/[ATCG]/){
        $data{total} ++ ;
        if(is_hetero($gt) eq 1){
            $data{hom}++;
        }else{
            $data{het}++;
        }
        $data{dbSNP} ++ if $rs ne '.';
        $data{TG} ++ if $tg ne '.' ;
        $data{locale}{$func} ++ if $func=~/\w/;
        $data{result}{$efunc} ++ if $efunc;
        $data{transfm_CDS}{$ref.'->'.$alt}++ if($func=~/^exonic/);# or $locate=~/[^_]UTR/
       # $data{transfm}{$ref.'->'.$alt}++;
    }elsif($ref=~/-/ or $alt=~/-/){
        $indel{total} ++ ;
        $indel{ins} ++ if $ref=~/-/;
        $indel{del} ++ if $alt=~/-/;
        if(is_hetero($gt) eq 1){
            $indel{hom}++
        }else{
            $indel{het}++
        }
        $indel{dbSNP} ++ if $rs ne '.';
        $indel{TG} ++ if $tg ne '.' ;
        $indel{locale}{$func} ++ if $func=~/\w/;
        $indel{result}{$efunc} ++ if $efunc;
    }
}
close($in_fh);

my ($total,$het,$hom,$het2hom,$dbsnp,$tg);
$data{total}=defined $data{total} ? $data{total} :0 ;
my $dbsnp_ratio=defined $data{dbSNP} ? 100*$data{dbSNP}/$data{total}: 0;
my $tg_ratio=defined $data{TG} ? 100*$data{TG}/$data{total} :0;
$dbsnp_ratio=sprintf("%.2f",$dbsnp_ratio);
$tg_ratio=sprintf("%.2f",$tg_ratio);
if(defined $total){
    $total .="\t".$data{total};
    $het .="\t".$data{het};
    $hom .="\t".$data{hom};
    $dbsnp .="\t".$dbsnp_ratio;
    $tg .= "\t".$tg_ratio;       
}else{
    $data{hom}= defined $data{hom}? $data{hom} :0;
    $data{het}= defined $data{het}? $data{het} :0;
    $total="Total_Num_SNP:\t".$data{total};
    $het="Heterozygotes:\t".$data{het};
    $hom="Homozygotes:\t".$data{hom};
    $het2hom=$data{hom}>0 ? "Het/Hom:\t".sprintf("%.2f",$data{het}/$data{hom}) : "Het/Hom:\tNA" ;
    $dbsnp="dbSNP(%):\t".$dbsnp_ratio;
    $tg="gnomAD_genome(%):\t".$tg_ratio;            
}

my ($total_id,$het_id,$hom_id,$het2hom_id,$dbsnp_id,$tg_id,$ins,$del);

$indel{total} = defined $indel{total} ?  $indel{total} : 0;
my $dbsnp_ratio_id=defined $indel{dbSNP} ? 100*$indel{dbSNP}/$indel{total} : 0;
my $tg_ratio_id=defined $indel{TG} ? 100*$indel{TG}/$indel{total} :0 ;
$dbsnp_ratio_id=sprintf("%.2f",$dbsnp_ratio_id);
$tg_ratio_id=sprintf("%.2f",$tg_ratio_id);
if(defined $total_id){
    $total_id .="\t".$indel{total};
    $ins .="\t".$indel{ins};
    $del .="\t".$indel{del};
    $het_id .="\t".$indel{het};
    $hom_id .="\t".$indel{hom};
    $dbsnp_id .="\t".$dbsnp_ratio_id;
    $tg_id .= "\t".$tg_ratio_id;       
}else{
    $total_id="Total_Num_INDEL:\t".$indel{total};
    $indel{ins}=defined $indel{ins} ? $indel{ins}:0;
    $indel{del}=defined $indel{del} ? $indel{del}:0;
    $indel{het}=defined $indel{het} ? $indel{het}:0;
    $indel{hom}=defined $indel{hom} ? $indel{hom}:0;
    $ins="Insertions:\t".$indel{ins};
    $del="Deletions:\t".$indel{del};
    $het_id="Heterozygotes:\t".$indel{het};
    $hom_id="Homozygotes:\t".$indel{hom};
    $het2hom_id=$indel{hom} > 0 ? "Het/Hom:\t".sprintf("%.2f",$indel{het}/$indel{hom}) : "Het/Hom:\tNA";
    $dbsnp_id="dbSNP(%):\t".$dbsnp_ratio_id;
    $tg_id="gnomAD_genome(%):\t".$tg_ratio_id;            
}

my($id_locations,$id_results, $locations,$results);   
$locations=hash2string($data{locale},0);
$results=hash2string($data{result},1);
$id_locations=hash2string($indel{locale},0);
$id_results=hash2string($indel{result},1);

for my $transfm(keys %{$data{transfm_CDS}}){
    if ($transfm eq 'A->G' or $transfm eq 'G->A'
            or $transfm eq 'T->C' or $transfm eq 'C->T'){
            $data{'Ti(CDS)'} += $data{transfm_CDS}{$transfm};
    }else{
        $data{'Tv(CDS)'} += $data{transfm_CDS}{$transfm};
    }       
}
#######Ti/Tv information collection total ###############
#for my $transfm(keys %{$data{transfm}}){
#    if ($transfm eq 'A->G' or $transfm eq 'G->A'
#            or $transfm eq 'T->C' or $transfm eq 'C->T'){
#            $data{Ti} += $data{transfm}{$transfm};
#    }else{
#        $data{Tv} += $data{transfm}{$transfm};
#    }       
#}
    
open my $w_fh,'>',$out_file or die "Cannot write file:$!";
my $title="Sample_Name:\t$sam";
print $w_fh join("\n",($title,$total,$het,$hom,$het2hom,$dbsnp,$tg)),"\n",$locations,$results;

######## print Ti Tv Ti/Tv information ############
#for my $i('Ti','Tv','Ti(CDS)','Tv_CDS'){
if (defined $data{'Ti(CDS)'}){
for my $i('Ti(CDS)','Tv(CDS)'){
    print $w_fh $i.':';
    print $w_fh "\t",$data{$i};
    print $w_fh "\n";
}   
}
#print $w_fh 'Ti/Tv:';
#print $w_fh "\t",sprintf "%.2f",$data{Ti}/$data{Tv};
#print $w_fh "\n";
if(defined $data{'Ti(CDS)'}){
    print $w_fh 'Ti/Tv(CDS):';
    print $w_fh "\t",sprintf "%.2f",$data{'Ti(CDS)'}/$data{'Tv(CDS)'};
    print $w_fh "\n";
}
print $w_fh "######INDEL######\n";
print $w_fh join("\n",($total_id,$dbsnp_id,$tg_id,$ins,$del,$het_id,$hom_id,$het2hom_id)),"\n",$id_locations,$id_results; 

sub is_hetero{
    my ($gt)=shift;
    my ($a,$b)=split /\//,$gt;
    if($a==$b){
        return 1;
    }else{
        return 0;
    }   
}

sub is_alt{
    my ($gt,$alts,$alt)=@_;
    my ($a,$b)=split /\//,$gt;
    my @base=split /,/,$alts;
    @base=(' ',@base);
    my $flag;
    for($a,$b){
        if($base[$_] eq $alt){
            return 1;
        }
    }
    return 0;   
}

sub hash2string{
    my ($hash,$gap)=@_;
    unless (defined $hash){
        return '';
    }
    my %hash=%{$hash};
    my $string='';
    my $pre=$gap? " ":"";
    for my $result(sort keys %hash){
        next if $result eq '.' or $result=~/ncRNA/ or $result=~/intronic/ or 
                $result=~/stream/ or  $result=~/intergenic/ or $result=~/UTR/ ;
        if(defined $string){
            $string .= $pre.$result.":";
        }else{
            $string= $pre.$result.":";
        }
        $hash{$result}=defined $hash{$result}?$hash{$result}:0;
        $string .="\t".$hash{$result};
        $string .="\n";
    }
    return $string;
}