问题描述
给出以下文件:
NW_022983499.1 RefSeq CDS 6883 7503 . + 0 ID=cds-XP_033376633.1
NW_022983500.1 RefSeq CDS 5353 5898 . + 0 ID=cds-XP_033376630.1
NW_022983500.1 RefSeq CDS 6033 7994 . + 0 ID=cds-XP_033376630.1
NW_022983502.1 RefSeq CDS 5391 5543 . + 0 ID=cds-XP_033376626.1
NW_022983502.1 RefSeq CDS 5591 5673 . + 0 ID=cds-XP_033376626.1
NW_022983502.1 RefSeq CDS 5782 5895 . + 1 ID=cds-XP_033376626.1
NW_022983502.1 RefSeq CDS 5937 6424 . + 1 ID=cds-XP_033376626.1
NW_022983502.1 RefSeq CDS 6478 6680 . + 2 ID=cds-XP_033376626.1
NW_022983502.1 RefSeq CDS 6739 6858 . + 0 ID=cds-XP_033376626.1
NW_022983502.1 RefSeq CDS 6926 7408 . + 0 ID=cds-XP_033376626.1
NW_022983504.1 RefSeq CDS 5478 5513 . - 0 ID=cds-XP_033376620.1
NW_022983504.1 RefSeq CDS 5353 5419 . - 0 ID=cds-XP_033376620.1
NW_022983504.1 RefSeq CDS 5161 5297 . - 2 ID=cds-XP_033376620.1
NW_022983504.1 RefSeq CDS 5059 5115 . - 0 ID=cds-XP_033376620.1
NW_022983508.1 RefSeq CDS 4415 5392 . - 1 ID=cds-XP_033376609.1
NW_022983508.1 RefSeq CDS 4215 4344 . - 1 ID=cds-XP_033376609.1
NW_022983512.1 RefSeq CDS 2650 2831 . + 0 ID=cds-XP_033376596.1
NW_022983512.1 RefSeq CDS 2890 3112 . + 1 ID=cds-XP_033376596.1
NW_022983512.1 RefSeq CDS 3163 3267 . + 0 ID=cds-XP_033376596.1
我想提取一组对应于第9列中ID的坐标(从低到高的数值),以获得以下文件:
NW_022983499.1 RefSeq CDS 6883 7503 . + 0 ID=cds-XP_033376633.1
NW_022983500.1 RefSeq CDS 5353 7994 . + 0 ID=cds-XP_033376630.1
NW_022983502.1 RefSeq CDS 5391 7408 . + 0 ID=cds-XP_033376626.1
NW_022983504.1 RefSeq CDS 5059 5513 . - 0 ID=cds-XP_033376620.1
NW_022983508.1 RefSeq CDS 4215 5392 . - 0 ID=cds-XP_033376609.1
NW_022983512.1 RefSeq CDS 2650 3267 . + 0 ID=cds-XP_033376596.1
请注意,对于ID=cds-XP_033376630.1
在第7列中具有正值的情况,我需要选择第2行第4列5353
和第3行第5列7994
的值
相反,如果第7列的值为负,如ID=cds-XP_033376620.1
,则逻辑取反,我需要选择第14行的第4列5059
和第11行的值第5栏5513
我对使用AWK(不是Perl或Python)来解决这种经典的生物信息学问题特别感兴趣,如果有人能指出正确的方向,我将不胜感激。
解决方法
$ cat tst.awk
$NF != prevKey {
if ( NR > 1 ) {
prt()
}
min = $4
max = $5
line = $0
prevKey = $NF
}
{
min = ($4 <= min ? $4 : min)
max = ($4 >= max ? $5 : max)
}
END { prt() }
function prt( orig) {
orig = $0
$0 = line
$4 = min
$5 = max
$8 = 0
print
$0 = orig
}
。
$ awk -f tst.awk file
NW_022983499.1 RefSeq CDS 6883 7503 . + 0 ID=cds-XP_033376633.1
NW_022983500.1 RefSeq CDS 5353 7994 . + 0 ID=cds-XP_033376630.1
NW_022983502.1 RefSeq CDS 5391 7408 . + 0 ID=cds-XP_033376626.1
NW_022983504.1 RefSeq CDS 5059 5513 . - 0 ID=cds-XP_033376620.1
NW_022983508.1 RefSeq CDS 4215 5392 . - 0 ID=cds-XP_033376609.1
NW_022983512.1 RefSeq CDS 2650 3267 . + 0 ID=cds-XP_033376596.1
,
$ awk 'p9!=$9{if(p0) print p0} !a[$9]++; {p9=$9; p0=$0} END{print p0}' file |
awk 'NR%2{k=($7=="+")?4:5; v=$k; next} {$k=v}1'
NW_022983499.1 RefSeq CDS 6883 7503 . + 0 ID=cds-XP_033376633.1
NW_022983500.1 RefSeq CDS 5353 7994 . + 0 ID=cds-XP_033376630.1
NW_022983502.1 RefSeq CDS 5391 7408 . + 0 ID=cds-XP_033376626.1
NW_022983504.1 RefSeq CDS 5059 5513 . - 0 ID=cds-XP_033376620.1
NW_022983508.1 RefSeq CDS 4215 5392 . - 1 ID=cds-XP_033376609.1
NW_022983512.1 RefSeq CDS 2650 3267 . + 0 ID=cds-XP_033376596.1
两个单独的脚本将简化逻辑,第一个脚本打印每个键的第一行和最后一行(如果存在一行,则重复)。第二个脚本根据符号选择正确的值。
,另一个awk
(还添加了$ 0的打印零,如注释所示)
> cat tst.awk
$9 == prev {
$keep = val
$8 = 0
row = $0
next
}
{
print row
prev = $9
$8 = 0
row = $0
keep = ( $7=="+"? 4: 5 )
val = $keep
}
END {
print row
}
输出:
> awk -f tst.awk file
NW_022983499.1 RefSeq CDS 6883 7503 . + 0 ID=cds-XP_033376633.1
NW_022983500.1 RefSeq CDS 5353 7994 . + 0 ID=cds-XP_033376630.1
NW_022983502.1 RefSeq CDS 5391 7408 . + 0 ID=cds-XP_033376626.1
NW_022983504.1 RefSeq CDS 5059 5513 . - 0 ID=cds-XP_033376620.1
NW_022983508.1 RefSeq CDS 4215 5392 . - 0 ID=cds-XP_033376609.1
NW_022983512.1 RefSeq CDS 2650 3267 . + 0 ID=cds-XP_033376596.1