#!/usr/bin/perl -T

##################################################################
#	Validation format for PCM1 submission                    #
#	Number of rows: 38 + 1 header row                      #
#	Number of columns: 9                                     #
##################################################################

use strict;
use warnings;
my $usage="

# This script is used for PCM1 challenge of CAGI-5.
# Please run this script before your submission to ensure the data format of submitted file is correct!

# Data format 
#     tab-seperated 9 columns file:
# 1. Nucleotide position: DNA coding change of the variant (e.g., c.G17A) on the NM_001315507 transcript.
# 2. Variant: protein change of the variant (e.g. p.G6D), on UniProtKB protein - Q15154 (PCM1_HUMAN)
# 3. p-value relative of change from MO: The probability that this variant is statistically different from MO.
# 4. Standard deviation: This defines the confidence of the prediction in column 3. Large SD means low confidence, while small SD means that the predictor is confident about the submitted prediction.
# 5.p-value of change from MO+WT: The probability that this variant is statistically different from MO+WT
# 6. Standard deviation: This defines the confidence of the prediction in column 5. Large SD means low confidence, while small SD means that the predictor is confident about the submitted prediction.
# 7. Functional effect: pathogenic (2), hypomorphic (1), or benign (0)
# 8. Confidence: in the functional effect assignment ranges 0.0 to 1.0. (1.0 implies total confidence in the assignment)
# 9. Comments: Optional brief comments based on the predictions	 
#  Leave a \"*\" if you have no predictions/commets.

Usage: perl $0 <submitted file>

Thank you for taking part in CAGI 5! Good luck!

";

die $usage if @ARGV!=1;

my $num_rows = 39; # including header
my $num_columns = 9;
my $header="Nucleotide_position\tVariant\tP-value_MO\tStandard_deviation\tP-value_MO+WT\tStandard_deviation\tFunctional_effect\tConfidence:\tComments";


my $filename = $ARGV[0];
#if ($filename =~ /^([A-Za-z\_\-\.0-9]+)$/){
if ($filename =~ /^([A-Za-z\-\_\.\/0-9]+)$/){
    $filename = $1;
} 
else {
    die "Only alphabet, numbers, \"_\", \"-\" and \".\" are allowed in the file name\n";
}

open(IFILE, $filename) or die "Cannot open input file: $!\n";
my @contents = <IFILE>;
chomp @contents;
close(IFILE);

my @errors = ();
push @errors, "The number of lines does not match the template (1 header + 38 data rows)!\n" if @contents != $num_rows;

my $h=shift @contents;
push @errors, "Header should be exact \"$header\".\n" if $h ne $header;

for(my $line=1;$line<=@contents;$line++){

    my @cols = split "\t",$contents[$line-1];
    push @errors,"The number of columns at line $line should be exact $num_columns. Please indicate the empty columns with '*'.\n" if @cols < $num_columns;
    push @errors,"The number of columns at line $line should be exact $num_columns. Tabs can only be used as column separators.\n" if @cols > $num_columns;

    #push @errors, "Invalid Nucleotide Position: $cols[0] at line $line. Nucleotide Position should be the same as in the test data.\n" unless $cols[0]=~/^[ACTG]{1}\d+[ACTG]{1}$/;
    #push @errors, "Invalid Variant: $cols[0] at line $line. Variant should be the same as in the test data.\n" unless $cols[1]=~/^[A-Z]{1}\d+[A-Z]{1}$/;  

   # unless($cols[1] eq "*"){
   # 	if($cols[1]=~/^[+-]?\d+\.?\d*$/){
   # 	    push @errors, "The 2nd column (relative activity) at line $line should be a real number; leave a '*' to indicate empty number.\n" if($cols[1]<0);
   # 	} else{
   # 	    push @errors, "The 2nd column (relative activity) at line $line should be a real number; leave a '*' to indicate empty number.\n"; 
   # 	}
   # }
 
    unless($cols[2] eq "*"){
        if($cols[2]=~/^[+-]?\d+\.?\d*$/){
            push @errors, "The 3rd column (p-value relative of change from MO) at line $line should be a number; leave a '*' to indicate empty number.\n" if($cols[3]>1 || $cols[3]<0);
        } else{
            push @errors, "The 3rd column (p-value relative of change from MO) at line $line should be a real number; leave a '*' to indicate empty number.\n"; 
        }
    }
    
    unless($cols[3] eq "*"){
        if($cols[3]=~/^[+-]?\d+\.?\d*$/){
            push @errors, "The 4th column (SD) at line $line should be a real number between 0 and 1; leave a '*' to indicate empty number.\n" if($cols[3]>1 || $cols[3]<0);
        } else{
            push @errors, "The 4th column (SD) at line $line should be a real number; leave a '*' to indicate empty number.\n"; 
        }
    }

    unless($cols[4] eq "*"){
        if($cols[4]=~/^[+-]?\d+\.?\d*$/){
            push @errors, "The 5th column (p-value of change from MO+WT) at line $line should be a number; leave a '*' to indicate empty number.\n" if($cols[3]>1 || $cols[3]<0);
        } else{
            push @errors, "The 5th column (p-value of change from MO+WT) at line $line should be a real number; leave a '*' to indicate empty number.\n"; 
        }
    }

    unless($cols[5] eq "*"){
        if($cols[5]=~/^[+-]?\d+\.?\d*$/){
            push @errors, "The 6th column (SD) at line $line should be a real number between 0 and 1; leave a '*' to indicate empty number.\n" if($cols[3]>1 || $cols[3]<0);
        } else{
            push @errors, "The 6th column (SD) at line $line should be a real number; leave a '*' to indicate empty number.\n"; 
        }
    }

    # unless($cols[6] eq "*"){
    #     if($cols[6]=~/^[+-]?\d+\.?\d*$/){
    #        push @errors, "The 7th column (p-value relative of change from MO) at line $line should be a real number; leave a '*' to indicate empty number.\n" if($cols[3]>1 || $cols[3]<0);
    #   } else{
    #      push @errors, "The 7th column (p-value relative of change from MO) at line $line should be a real number; leave a '*' to indicate empty number.\n"; 
    # }
    #}
    
    unless($cols[7] eq "*"){
        if($cols[7]=~/^[+-]?\d+\.?\d*$/){
            push @errors, "The 8th column (Confidence) at line $line should be a number; leave a '*' to indicate empty number.\n" if($cols[3]>1 || $cols[3]<0);
        } else{
            push @errors, "The 8th column (Confidence) at line $line should be a real number; leave a '*' to indicate empty number.\n"; 
        }
    }


    unless($cols[8] eq "*"){
        if($cols[8]=~/^\w+$/){
            push @errors, "The 9th column (Comments) at line $line should be any text; leave a '*' to indicate empty.\n" if($cols[5]=~/^\W+$/);  #non word characters will error
        }
    }

}

if (@errors > 0) {
    print "Validation failed with errors:\n";
    print @errors;
    exit -1;
} 
else {
    print "Congratulations! Your submitted file has a validated format.\n";
    if ((-x "/usr/bin/md5sum") && (-x "/usr/bin/cut")) {
	print "MD5 hash for your file is ";
	$ENV{PATH} = "";
	system ("/usr/bin/md5sum $filename | /usr/bin/cut -c 1-32");
	#print "\n";
	if ((-r "/data/ajitha/checks/cagi.txt") && (-x "/bin/cat")) {
	    print "Receipt code for your file is ";
	    system ("/bin/cat /data/ajitha/checks/cagi.txt $filename | /usr/bin/md5sum | /usr/bin/cut -c 1-32");
	}
    }
    exit 0;
}