使用软件 CNVnator_v0.3.2

过滤脚本 cnvnator_filter.pl

CNV输出结果

CNV_type        coordinates     CNV_size        normalized_RD   p-val1  p-val2  p-val3  p-val4  q0
duplication     chr10:1-9590600 9.5906e+06      3.73714e+07     0       0       0       0       0.0087298
deletion        chr10:9590601-9598000   7400    0       193986  1.51989e-13     265833  1.59372e-07
     0
duplication     chr10:9598001-20649900  1.10519e+07     3.11348e+07     0       0       0       0
       0.0164661
deletion        chr10:20649901-20666600 16700   2.04526e+07     0.107276        1.53471e-41     1.16667 1.60926e-35     0.0105263
duplication     chr10:20666601-58953000 3.82864e+07     2.68946e+07     0       0       0       0
       0.0687549
deletion        chr10:58956801-58966800 10000   0       143550  2.26481e-21     179437  2.37483e-15
     0
deletion        chr10:58967101-58971500 4400    0       326250  0.000163197     598125  171.125 1
duplication     chr10:58971501-67192000 8.2205e+06      2.50849e+07     0       0       0       0
       0.00724827

过滤后

CNV_type        coordinates     CNV_size        normalized_RD   p-val1  p-val2  p-val3  p-val4  q0
deletion        chr6:17969101-17990800  21700   0.0938754       7.34436e-12     1.50881e+09     1.0517e-10      1.60098e+09     0
deletion        chr6:51409801-51432200  22400   0.130833        1.36433e-05     2.85248e+09     0.00071419      2.85413e+09     0.2
deletion        chr6:62722501-62745900  23400   0.155307        0.00782878      2.87083e+09     54829.5 7.85916e-67     0
deletion        chr6:63945701-63964200  18500   0.158415        0.0224522       2.8557e+09      0.60604 2.85735e+09     0.5
deletion        chr6:84412301-84430800  18500   0.158415        0.0224522       2.8557e+09      0.60604 2.85735e+09     0.5
deletion        chr6:94104601-94138600  34000   0.11393 3.51557e-09     2.87096e+09     34881.6 2.93187e-104    0
deletion        chr6:95677601-95830900  153300  0.175024        2.02724e-10     2.871e+09       1.05335e-12     2.871e+09       0
deletion        chr6:101751601-101764300        12700   0.157061        0.0380917       1.82699e+09
     3.6824  1.96178e+09     0.5
deletion        chr6:128498001-128514200        16200   0.138074        0.000294307     2.36381e+09
     86017.5 2.08126e-41     0
deletion        chr6:157641201-157696200        55000   0.132152        2.99676e-06     2.871e+09
       19396   2.07431e-178    0
deletion        chr6:167942101-168043600        101500  0.0540112       1.57017e-12     2.871e+09
       8919.17 0       0
deletion        chr6:171052901-171115100        62200   0.150971        0.000636157     2.871e+09
       16651.8 7.83291e-204    0.333333
deletion        chrM:8801-16600 7800    0.590908        2.54587e-08     2.42617e+09     0.000434635
     2.53319e+09     0.00227964

过滤脚本

#!/usr/bin/perl -w
use strict;

die "Usage:\n\tperl filterCNV.pl <CNVs> \n" unless @ARGV == 1;

my $cnv = shift;

open CNV,"<$cnv" || die "$!";
my @pl = "";

while(<CNV>) {
        chomp;
        if ($_ =~ /normalized_RD/)
        {
                print "$_\n";
                next;
        }
        @pl = split /\t/,$_;

        if($pl[3] ==0 || $pl[3] >4 ||$pl[8] == -1 || $pl[8] > 0.5 || $pl[2] < 1000) {
                next;   
        }else {
                print $_ . "\n";        
        }
}

close CNV;

结语:

CNV的可靠性不高,一般人可以检测出1k~2k个,没有较好的验证方法(exome SNP还可以用芯片来评估)

所以建议多用几个软件来call CNV,另外可以设置上面的q0 参数,到0.1,甚至0.05(上面脚本中的pl[8])

版权声明:本文为leezx原创文章,遵循 CC 4.0 BY-SA 版权协议,转载请附上原文出处链接和本声明。
本文链接:https://www.cnblogs.com/leezx/p/6650071.html