Linux的日常·单行数据转fasta

原谅我真的不知道怎么取这次的题目……

一个朋友跑来问我， 怎么使用linux将一个名称和序列单行存放的数据转成fasta格式。

还是直接上图比较直观：

原始的文件是这样的：

名称1 tab分隔 名称2 tab分隔 名称3 tab分隔 序列 tab分隔 null

转成：

>名称1|名称2|名称3
序列

好，那第一步先把tab分隔变成分行：

$ cat temp |sed 's/\t/\n/g'|head -8
MetaCYC
G-12237
Q7X281
MKVACIGAGPGGLFFATLLKRSRPDAEVVVFERNRPDDTFGFGVVFSDATLDAIDAADPVLSEALEKHGRHWDDIEVRVHGERERVGGMGMAAVVRKTLLSLLQERARAEGVQMRFQDEVRDPAELDDFDLVVVCDGANSRFRTLFADDFGPTAEVASAKFIWFGTTYMFDGLTFVHQDGPHGVFAAHAYPISDSLSTFIVETDADSWARAGLDAFDPATPLGMSDEKTKSYLEDLFRAQIDGHPLVGNNSRWANFATRRARSWRSGKWVLLGDAAHTAHFSVGSGTKMAMEDAVALAETLGEASRSVPEALDLYEERRRPKVERIQNSARPSLSWWEHFGRYVRSFDAPTQFAFHFLTRSIPRGKLAVRDAAYVDRVDGWWLRHHEAGPLKTPFRVGPYRLPTRRVTVGDDLLTGTDGTGIPMVPFSGQPFGAGVWIDAPDTEEGLPLALDQVRETAEAGVLLVGVRGGTALTRVLVAEEARLAHSLPAAIVGAYDDDTATTLVLSGRADLVGGTK
null
MetaCYC
EG10460
P05719

稍微解释一下：

• s ：是替换字符串的命令
• \t ：tab分隔符
• \n ：换行符
• g ：全局替换。不加的话只替换第一个，加上就是替换所有的。

接下来用 grep -v 把null这行去掉：

$ cat temp |sed 's/\t/\n/g'|grep -v "null"|head -8
MetaCYC
G-12237
Q7X281
MKVACIGAGPGGLFFATLLKRSRPDAEVVVFERNRPDDTFGFGVVFSDATLDAIDAADPVLSEALEKHGRHWDDIEVRVHGERERVGGMGMAAVVRKTLLSLLQERARAEGVQMRFQDEVRDPAELDDFDLVVVCDGANSRFRTLFADDFGPTAEVASAKFIWFGTTYMFDGLTFVHQDGPHGVFAAHAYPISDSLSTFIVETDADSWARAGLDAFDPATPLGMSDEKTKSYLEDLFRAQIDGHPLVGNNSRWANFATRRARSWRSGKWVLLGDAAHTAHFSVGSGTKMAMEDAVALAETLGEASRSVPEALDLYEERRRPKVERIQNSARPSLSWWEHFGRYVRSFDAPTQFAFHFLTRSIPRGKLAVRDAAYVDRVDGWWLRHHEAGPLKTPFRVGPYRLPTRRVTVGDDLLTGTDGTGIPMVPFSGQPFGAGVWIDAPDTEEGLPLALDQVRETAEAGVLLVGVRGGTALTRVLVAEEARLAHSLPAAIVGAYDDDTATTLVLSGRADLVGGTK
MetaCYC
EG10460
P05719
MSAGKLPEGWVIAPVSTVTTLIRGVTYKKEQAINYLKDDYLPLIRANNIQNGKFDTTDLVFVPKNLVKESQKISPEDIVIAMSSGSKSVVGKSAHQHLPFECSFGAFCGVLRPEKLIFSGFIAHFTKSSLYRNKISSLSAGANINNIKPASFDLINIPIPPLAEQKIIAEKLDTLLAQVDSTKARFEQIPQILKRFRQAVLGGAVNGKLTEKWRNFEPQHSVFKKLNFESILTELRNGLSSKPNESGVGHPILRISSVRAGHVDQNDIRFLECSESELNRHKLQDGDLLFTRYNGSLEFVGVCGLLKKLQHQNLLYPDKLIRARLTKDALPEYIEIFFSSPSARNAMMNCVKTTSGQKGISGKDIKSQVVLLPPVKEQAEIVRRVEQLFAYADTIEKQVNNALARVNNLTQSILAKAFRGELTAQWRAENPDLISGENSAAALLEKIKAERAASGGKKASRKKS

grep -v "null" 是去除含有"null"字符串的行。也可以写的更严谨一些： grep -v "^null$" 。

• ^ 表示行首
• $ 表示行尾

接下来，就是把前面3行变成一行，并且 | 分隔。

$ cat temp |sed 's/\t/\n/g'|grep -v "null"|awk '{if(NR%4!=0){printf "|%s",$0}else{printf "\n" $0 "\n"}}'|head -4
|MetaCYC|G-12237|Q7X281
MKVACIGAGPGGLFFATLLKRSRPDAEVVVFERNRPDDTFGFGVVFSDATLDAIDAADPVLSEALEKHGRHWDDIEVRVHGERERVGGMGMAAVVRKTLLSLLQERARAEGVQMRFQDEVRDPAELDDFDLVVVCDGANSRFRTLFADDFGPTAEVASAKFIWFGTTYMFDGLTFVHQDGPHGVFAAHAYPISDSLSTFIVETDADSWARAGLDAFDPATPLGMSDEKTKSYLEDLFRAQIDGHPLVGNNSRWANFATRRARSWRSGKWVLLGDAAHTAHFSVGSGTKMAMEDAVALAETLGEASRSVPEALDLYEERRRPKVERIQNSARPSLSWWEHFGRYVRSFDAPTQFAFHFLTRSIPRGKLAVRDAAYVDRVDGWWLRHHEAGPLKTPFRVGPYRLPTRRVTVGDDLLTGTDGTGIPMVPFSGQPFGAGVWIDAPDTEEGLPLALDQVRETAEAGVLLVGVRGGTALTRVLVAEEARLAHSLPAAIVGAYDDDTATTLVLSGRADLVGGTK
|MetaCYC|EG10460|P05719
MSAGKLPEGWVIAPVSTVTTLIRGVTYKKEQAINYLKDDYLPLIRANNIQNGKFDTTDLVFVPKNLVKESQKISPEDIVIAMSSGSKSVVGKSAHQHLPFECSFGAFCGVLRPEKLIFSGFIAHFTKSSLYRNKISSLSAGANINNIKPASFDLINIPIPPLAEQKIIAEKLDTLLAQVDSTKARFEQIPQILKRFRQAVLGGAVNGKLTEKWRNFEPQHSVFKKLNFESILTELRNGLSSKPNESGVGHPILRISSVRAGHVDQNDIRFLECSESELNRHKLQDGDLLFTRYNGSLEFVGVCGLLKKLQHQNLLYPDKLIRARLTKDALPEYIEIFFSSPSARNAMMNCVKTTSGQKGISGKDIKSQVVLLPPVKEQAEIVRRVEQLFAYADTIEKQVNNALARVNNLTQSILAKAFRGELTAQWRAENPDLISGENSAAALLEKIKAERAASGGKKASRKKS

awk '{if(NR%4!=0){printf "|%s",$0}else{printf "\n" $0 "\n"}}' 的含义是，如果行号（NR）取余不等于0（即非4倍数的行），那么打印 | 和整行内容（$0），否则就打印换行符+整行+换行符。

最后就比较简单了。把 > 加上：

$ cat temp |sed 's/\t/\n/g'|grep -v "null"|awk '{if(NR%4!=0){printf "|%s",$0}else{printf "\n" $0 "\n"}}'|sed 's/|MetaCYC/>MetaCYC/'|head -4
>MetaCYC|G-12237|Q7X281
MKVACIGAGPGGLFFATLLKRSRPDAEVVVFERNRPDDTFGFGVVFSDATLDAIDAADPVLSEALEKHGRHWDDIEVRVHGERERVGGMGMAAVVRKTLLSLLQERARAEGVQMRFQDEVRDPAELDDFDLVVVCDGANSRFRTLFADDFGPTAEVASAKFIWFGTTYMFDGLTFVHQDGPHGVFAAHAYPISDSLSTFIVETDADSWARAGLDAFDPATPLGMSDEKTKSYLEDLFRAQIDGHPLVGNNSRWANFATRRARSWRSGKWVLLGDAAHTAHFSVGSGTKMAMEDAVALAETLGEASRSVPEALDLYEERRRPKVERIQNSARPSLSWWEHFGRYVRSFDAPTQFAFHFLTRSIPRGKLAVRDAAYVDRVDGWWLRHHEAGPLKTPFRVGPYRLPTRRVTVGDDLLTGTDGTGIPMVPFSGQPFGAGVWIDAPDTEEGLPLALDQVRETAEAGVLLVGVRGGTALTRVLVAEEARLAHSLPAAIVGAYDDDTATTLVLSGRADLVGGTK
>MetaCYC|EG10460|P05719
MSAGKLPEGWVIAPVSTVTTLIRGVTYKKEQAINYLKDDYLPLIRANNIQNGKFDTTDLVFVPKNLVKESQKISPEDIVIAMSSGSKSVVGKSAHQHLPFECSFGAFCGVLRPEKLIFSGFIAHFTKSSLYRNKISSLSAGANINNIKPASFDLINIPIPPLAEQKIIAEKLDTLLAQVDSTKARFEQIPQILKRFRQAVLGGAVNGKLTEKWRNFEPQHSVFKKLNFESILTELRNGLSSKPNESGVGHPILRISSVRAGHVDQNDIRFLECSESELNRHKLQDGDLLFTRYNGSLEFVGVCGLLKKLQHQNLLYPDKLIRARLTKDALPEYIEIFFSSPSARNAMMNCVKTTSGQKGISGKDIKSQVVLLPPVKEQAEIVRRVEQLFAYADTIEKQVNNALARVNNLTQSILAKAFRGELTAQWRAENPDLISGENSAAALLEKIKAERAASGGKKASRKKS

到这里其实已经结束了。但是我还是感觉写的太罗嗦了。于是换了一种解决方法：

$ cat temp |sed 's/\t/\n/g'|grep -v "null"|sed 'N;N;s/\n/|/g;s/^/>/;N'|head -4
>MetaCYC|G-12237|Q7X281
MKVACIGAGPGGLFFATLLKRSRPDAEVVVFERNRPDDTFGFGVVFSDATLDAIDAADPVLSEALEKHGRHWDDIEVRVHGERERVGGMGMAAVVRKTLLSLLQERARAEGVQMRFQDEVRDPAELDDFDLVVVCDGANSRFRTLFADDFGPTAEVASAKFIWFGTTYMFDGLTFVHQDGPHGVFAAHAYPISDSLSTFIVETDADSWARAGLDAFDPATPLGMSDEKTKSYLEDLFRAQIDGHPLVGNNSRWANFATRRARSWRSGKWVLLGDAAHTAHFSVGSGTKMAMEDAVALAETLGEASRSVPEALDLYEERRRPKVERIQNSARPSLSWWEHFGRYVRSFDAPTQFAFHFLTRSIPRGKLAVRDAAYVDRVDGWWLRHHEAGPLKTPFRVGPYRLPTRRVTVGDDLLTGTDGTGIPMVPFSGQPFGAGVWIDAPDTEEGLPLALDQVRETAEAGVLLVGVRGGTALTRVLVAEEARLAHSLPAAIVGAYDDDTATTLVLSGRADLVGGTK
>MetaCYC|EG10460|P05719
MSAGKLPEGWVIAPVSTVTTLIRGVTYKKEQAINYLKDDYLPLIRANNIQNGKFDTTDLVFVPKNLVKESQKISPEDIVIAMSSGSKSVVGKSAHQHLPFECSFGAFCGVLRPEKLIFSGFIAHFTKSSLYRNKISSLSAGANINNIKPASFDLINIPIPPLAEQKIIAEKLDTLLAQVDSTKARFEQIPQILKRFRQAVLGGAVNGKLTEKWRNFEPQHSVFKKLNFESILTELRNGLSSKPNESGVGHPILRISSVRAGHVDQNDIRFLECSESELNRHKLQDGDLLFTRYNGSLEFVGVCGLLKKLQHQNLLYPDKLIRARLTKDALPEYIEIFFSSPSARNAMMNCVKTTSGQKGISGKDIKSQVVLLPPVKEQAEIVRRVEQLFAYADTIEKQVNNALARVNNLTQSILAKAFRGELTAQWRAENPDLISGENSAAALLEKIKAERAASGGKKASRKKS

看上去没多少人话了。。。。显得很专业的样子。

sed 'N;N;s/\n/|/g;s/^/>/;N' 是什么意思呢？

一个N相当于把当前行（MetaCYC）和下一行（G-12237）读取到模式空间中（就是某个地方存起来），这时候模式空间就有2行内容（MetaCYC换行符G-12237），且用换行符隔开。再加一个N，就是再下一行（第3行）也加进去。里面就存了3行内容了（MetaCYC换行符G-12237换行符Q7X281）。后边 s/\n/|/g 就是将换行符替换成 | （MetaCYC|G-12237|Q7X281），而 s/^/>/ 就是在前面处理的内容前面加 > （>MetaCYC|G-12237|Q7X281）。最后一个N就是把下一行放到这个处理好的行的后边。

好了~~今天到这。

放上本文的测试数据：

链接：https://pan.baidu.com/s/1w31KXlvyGOJqeS0dOfYsjA
提取码：rpag

如果觉得不错，点个“在看”哦~~