NGS基礎(chǔ) - GTF/GFF文件格式解讀和轉(zhuǎn)換這篇文章有讀者留言想要提取外顯子，內(nèi)含子，啟動(dòng)子，基因體，非編碼區(qū)，編碼區(qū)，TSS上游1500,TSS下游500得序列。下面我們就來示范如何提取這些序列。
NGS基礎(chǔ) - 參考基因組和基因注釋文件提到了如何下載對應(yīng)得基因組序列和基因注釋文件。
假如我們已經(jīng)拿到了基因組序列文件GRCh38.fa和基因注釋文件GRCh38.gtf，也可從文后鏈接獲取。
查看下文件內(nèi)容和格式基因組序列文件為FASTA格式，查看命令和內(nèi)容如下（測試文件，只有1條染色體）：
# 查看前10行，每行查看前40個(gè)字符# FASTA序列一般比較長，查看前面一部分字符是一個(gè)常用得方式head GRCh38.fa | cut -c 1-40>chr20NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN
基因注釋文件為GTF格式，只看前6列信息(第三列包含了不同得元件注釋）
cut -f 1-6 GRCh38.gtf | headchr20 ensembl_havana gene 87250 97094 .chr20 havana transcript 87250 97094 .chr20 havana exon 87250 87359 .chr20 havana exon 96005 97094 .chr20 ensembl_havana transcript 87710 96533 .chr20 ensembl_havana exon 87710 87767 .chr20 ensembl_havana CDS 87710 87767 .chr20 ensembl_havana start_codon 87710 87712 .chr20 ensembl_havana exon 96005 96533 .chr20 ensembl_havana CDS 96005 96414 .安裝提取工具gffread
這里用到了gffread (github/gpertea/gffread)，安裝方式如下 (若不理解，見這個(gè)為生信學(xué)習(xí)打造得開源Linux教程真香得軟件安裝部分)：
git clone github/gpertea/gffreadcd gffreadmake release提取轉(zhuǎn)錄本序列、CDS和蛋白序列
gffread -h可以參考所有可用參數(shù)，如果有特殊情況需要考慮得，還需配合其它參數(shù)使用。
1.獲取轉(zhuǎn)錄本序列
gffread GRCh38.gtf -g GRCh38.fa -w GRCh38.transcripts.fa
內(nèi)容如下：
head GRCh38.transcripts.fa>ENST00000608838ACAGGAATTCATATCGGGGTGATCACTCAGAAGAAAAGGTGAATACCGGATGTTGTAAGCTATTGAACTGCCACAAGTGATATCTTTACACACCATTCTGCTGTCATTGGGTAGCTTTGAACCCCAAAAATGTTGGAAGAATAATGTAGGACATTGCAGAAGACGATGTTTAGATACTGAAAGGTACATACTTCTTTGTAGGAACAAGCTATCATGCTGCATTTCTATAATATCACATGAATATACTCGACGACCAGCATTTCCTGTGATTCACCTAGAG
2.獲取CDS序列
# 獲取CDS序列g(shù)ffread GRCh38.gtf -g GRCh38.fa -x GRCh38.cds.fa
內(nèi)容如下
head GRCh38.cds.fa>ENST00000382410ATGAATATCCTGATGCTGACCTTCATTATCTGTGGGTTGCTAACTCGGGTGACCAAAGGTAGCTTTGAACCCCAAAAATGTTGGAAGAATAATGTAGGACATTGCAGAAGACGATGTTTAGATACTGAAAGGTACATACTTCTTTGTAGGAACAAGCTATCATGCTGCATTTCTATAATATCACATGAATATACTCGACGACCAGCATTTCCTGTGATTCACCTAGAGGATATAACATTGGATTATAGTGATGTGGACTCTTTTACTGGTTCCCCAGTATCTATGTTGAATGATCTGATAACATTTGACACAACTAAATTTGGAGAAACCATGACACCTGAGACCAATACTCCTGAGACTACTATGCCACCATCTGAGGCCACTACTCCCGAGACTACTATGCCACCATCTGAGACTGCTACTTCCGAGACTATGCCACCACCTTCTCAGACAGCTCTTACTCATAATTAA>ENST00000382398ATGAAGTCCCTACTGTTCACCCTTGCAGTTTTTATGCTCCTGGCCCAATTGGTCTCAGGTAATTGGTATG
3.獲取蛋白序列
# 獲取蛋白序列g(shù)ffread GRCh38.gtf -g GRCh38.fa -y GRCh38.protein.fa
內(nèi)容如下
head GRCh38.protein.fa>ENST00000382410MNILMLTFIICGLLTRVTKGSFEPQKCWKNNVGHCRRRCLDTERYILLCRNKLSCCISIISHEYTRRPAFPVIHLEDITLDYSDVDSFTGSPVSMLNDLITFDTTKFGETMTPETNTPETTMPPSEATTPETTMPPSETATSETMPPPSQTALTHN>ENST00000382398MKSLLFTLAVFMLLAQLVSGNWYVKKCLNDVGICKKKCKPEEMHVKNGWAMCGKQRDCCVPADRRANYPVFCVQTKTTRISTVTATTATTTLMMTTASMSSMAPTPVSPTG>ENST00000382388MGLFMIIAILLFQKPTVTEQLKKCWNNYVQGHCRKICRVNEVPEALCENGRYCCLNIKELEACKKITKPPRPKPATLALTLQDYVTIIENFPSLKTQST解析GTF文件得結(jié)構(gòu)
針對本GTF，對于gene元件，基因名字 (Gene symbol)在第14列。
head -n 1 GRCh38.gtf | sed 's/"/\t/g' | tr '\t' '\n' | sed = | sed 'N;s/\n/\t/'1 chr202 ensembl_havana3 gene4 872505 970946 .7 +8 .9 gene_id 10 ENSG0000017859111 ; gene_version 12 613 ; gene_name 14 DEFB12515 ; gene_source 16 ensembl_havana17 ; gene_biotype 18 protein_coding19 ;
針對本GTF，對于transcript元件，基因名字 (Gene symbol)在第18列。
sed -n '2p' GRCh38.gtf | sed 's/"/\t/g' | tr '\t' '\n' | sed = | sed 'N;s/\n/\t/'1 chr202 havana3 transcript4 872505 970946 .7 +8 .9 gene_id 10 ENSG0000017859111 ; gene_version 12 613 ; transcript_id 14 ENST0000060883815 ; transcript_version 16 117 ; gene_name 18 DEFB12519 ; gene_source 20 ensembl_havana21 ; gene_biotype 22 protein_coding23 ; transcript_name 24 DEFB125-20225 ; transcript_source 26 havana27 ; transcript_biotype 28 processed_transcript29 ; transcript_support_level 30 231 ;
這個(gè)查看信息在哪一列是很常用得檢查文件結(jié)構(gòu)提取對應(yīng)信息得方式，簡化為一個(gè)腳本checkCol.sh
檢查某個(gè)文件得指定行（默認(rèn)為第壹行）
checkCol.sh -f GRCh38.gtf1 chr202 ensembl_havana3 gene4 872505 970946 .7 +8 .9 gene_id "ENSG00000178591"; gene_version "6"; gene_name "DEFB125"; gene_source "ensembl_havana"; gene_biotype "protein_coding";
檢查標(biāo)準(zhǔn)輸入得第壹行
sed 's/"/\t/g' GRCh38.gtf | checkCol.sh -f -1 chr202 ensembl_havana3 gene4 872505 970946 .7 +8 .9 gene_id 10 ENSG0000017859111 ; gene_version 12 613 ; gene_name 14 DEFB12515 ; gene_source 16 ensembl_havana17 ; gene_biotype 18 protein_coding19 ;提取基因啟動(dòng)子序列
首先確定啟動(dòng)子區(qū)域，這里定義轉(zhuǎn)錄起始位點(diǎn)上游1000 bp和下游500 bp為啟動(dòng)子區(qū)域。
sed 's/"/\t/g' GRCh38.gtf | awk 'BEGIN{OFS=FS="\t"}{if($3=="gene") {if($7=="+") {start=$4-1000; end=$4+500;} else {if($7=="-") start=$5-500; end=$5+1000; } if(start<0) start=0; print $1,start,end,$14,$10,$7;}}' >GRCh38.promoter.bed
啟動(dòng)子區(qū)域如下 （這個(gè)bed文件也可以用于ChIP-seq類型得數(shù)據(jù)分析確定peak是否在啟動(dòng)子區(qū)域）
head GRCh38.promoter.bedchr20 86250 87750 DEFB125 ENSG00000178591 +chr20 141369 142869 DEFB126 ENSG00000125788 +chr20 156470 157970 DEFB127 ENSG00000088782 +chr20 189181 190681 DEFB128 ENSG00000185982 -chr20 226258 227758 DEFB129 ENSG00000125903 +chr20 256736 258236 DEFB132 ENSG00000186458 +chr20 266186 267686 AL034548.1 ENSG00000272874 +chr20 290278 291778 C20orf96 ENSG00000196476 -chr20 295968 297468 ZCCHC3 ENSG00000247315 +chr20 347724 349224 NRSN2-AS1 ENSG00000225377 -
然后提取序列。這里用到了bedtools工具，自家有提供編譯好得二進(jìn)制文件，下載下來即可使用。
# -name: 輸出基因名字（bed文件得第四列）# -s: 考慮到正反鏈（對于啟動(dòng)子區(qū)域，是否考慮鏈得信息關(guān)系不太大）bedtools getfasta -name -s -fi GRCh38.fa -bed GRCh38.promoter.bed >GRCh38.promoter.fa
序列信息如下：
head GRCh38.promoter.fa | cut -c 1-60>DEFB125::chr20:86250-87750(+)ATAATTTGAAGTGAGGTAATGTGATTCCTCTAGTTTTGTTCTTTTTGCTTAGGATGGCTT>DEFB126::chr20:141369-142869(+)AATATTCAAGAGAATGCCAAGAAAGCTACAAGAACAAATAGCAGGTCAGTCGTTGCCTGG>DEFB127::chr20:156470-157970(+)ATATCCGTCACCTCAAACATTTATCATTTGTATTGGGAACATTCAAAATCCTCTCTTCTA>DEFB128::chr20:189181-190681(-)AAAAAAGAAAAAGAACTCCAAGTCTAATAAGACCAGAGACCTGCCCTTTATGGGTCTGCA>DEFB129::chr20:226258-227758(+)GAGTGGAAGGTGGGAGGAGGGAGAGGATGAGGAAAAATAACTAATGGACACTAGGCTTAA
如果不想要坐標(biāo)信息，可對序列名字做一下簡化
cut -d ':' -f 1 GRCh38.promoter.fa >GRCh38.promoter.simplename.fahead GRCh38.promoter.simplename.fa | cut -c 1-60>DEFB125ATAATTTGAAGTGAGGTAATGTGATTCCTCTAGTTTTGTTCTTTTTGCTTAGGATGGCTT>DEFB126AATATTCAAGAGAATGCCAAGAAAGCTACAAGAACAAATAGCAGGTCAGTCGTTGCCTGG>DEFB127ATATCCGTCACCTCAAACATTTATCATTTGTATTGGGAACATTCAAAATCCTCTCTTCTA>DEFB128AAAAAAGAAAAAGAACTCCAAGTCTAATAAGACCAGAGACCTGCCCTTTATGGGTCTGCA>DEFB129GAGTGGAAGGTGGGAGGAGGGAGAGGATGAGGAAAAATAACTAATGGACACTAGGCTTAA提取基因序列
提取基因序列得操作也類似于提取啟動(dòng)子序列。這里要注意GFF文件得序列位置是從1開始，而bed文件得位置是從0開始，前閉后開，所以要對序列得起始位置進(jìn)行-1得操作。
type="gene"sed 's/"/\t/g' GRCh38.gtf | awk -v type="${type}" 'BEGIN{OFS=FS="\t"}{if($3==type) {print $1,$4-1,$5,$14,".",$7}}' >GRCh38.gene.bedhead GRCh38.gene.bedchr20 87249 97094 DEFB125 . +chr20 142368 145751 DEFB126 . +chr20 157469 159163 DEFB127 . +chr20 187852 189681 DEFB128 . -chr20 227257 229886 DEFB129 . +chr20 257735 261096 DEFB132 . +
提取基因序列
bedtools getfasta -name -s -fi GRCh38.fa -bed GRCh38.gene.bed >GRCh38.gene.fa# 查看序列head GRCh38.gene.fa | cut -c 1-60>DEFB125::chr20:87249-97094(+)ACAGGAATTCATATCGGGGTGATCACTCAGAAGAAAAGGTGAATACCGGATGTTGTAAGC>DEFB126::chr20:142368-145751(+)GCCATACACTTCAGCAGAGTTTGCAACTTCTCTTCTAAGTCTTTATCCTTCCCCCAAGGC>DEFB127::chr20:157469-159163(+)CTCTGAGGAAGGTAGCATAGTGTGCAGTTCACTGGACCAAAAGCTTTGGCTGCACCTCTT>DEFB128::chr20:187852-189681(-)GGCACACAGACCACTGGACAAAGTTCTGCTGCCTCTTTCTCTTGGGAAGTCTGTAAATAT提取非編碼RNA得序列
在GTF文件中有轉(zhuǎn)錄本類型得注釋，包含下面這些注釋類型
ntisense_RNAlincRNAmiRNAmisc_RNAprocessed_pseudogeneprocessed_transcriptprotein_codingrRNAscaRNAsense_intronicsense_overlappingsnoRNAsnRNATECtranscribed_processed_pseudogenetranscribed_unitary_pseudogenetranscribed_unprocessed_pseudogeneunitary_pseudogeneunprocessed_pseudogene
我們只篩選lincRNA
grep 'transcript_biotype "lincRNA"' GRCh38.gtf >GRCh38.lincRNA.gtfgffread GRCh38.lincRNA.gtf -g GRCh38.fa -w GRCh38.lincRNA.fahead GRCh38.lincRNA.fa | cut -c 1-60>ENST00000608495GTCGCACGCGCTGGCCAAACGGGCGCACCAGACACTTTTCAGGGCCCTGCCAAAGACCTCCTGGCGTCCCAGACACAAGAGATCCAGGCCAAGACTCACACTTCACAAGATACACAGACAGGAACAGGAAATTCCATGAAACTTCCATTTACCCAATTAGCCGGACTCACTGAGCCCCAGTCAACCAACTCCTACTAAAATTAAAAAGTAATGTGTGGTATAGATTGGAATAATAGACATAAACGATGGGAGGCGGAGAGGGGTGAGGGTTGAAAAATTACCTATTGGGTGCAACATTCAAATGGGGCACTAGAAGCCCACTCCACCACTATGCAATATATGTATTTGTACCCCGTAAAT提取一個(gè)個(gè)外顯子序列
獲取外顯子得坐標(biāo)
type="exon"sed 's/"/\t/g' GRCh38.gtf | awk -v type="${type}" 'BEGIN{OFS=FS="\t"}{if($3==type) {print $1,$4-1,$5,$14,$20,$7}}' >GRCh38.exon.bed# 查看文件內(nèi)容head GRCh38.exon.bedchr20 87249 87359 ENST00000608838 DEFB125 +chr20 96004 97094 ENST00000608838 DEFB125 +chr20 87709 87767 ENST00000382410 DEFB125 +chr20 96004 96533 ENST00000382410 DEFB125 +chr20 142368 142686 ENST00000382398 DEFB126 +chr20 145414 145751 ENST00000382398 DEFB126 +chr20 142633 142686 ENST00000542572 DEFB126 +chr20 145414 145488 ENST00000542572 DEFB126 +chr20 145578 145749 ENST00000542572 DEFB126 +chr20 157469 157593 ENST00000382388 DEFB127 +
提取序列
# -name: 輸出基因名字（bed文件得第四列）# -s: 考慮到正反鏈（對于啟動(dòng)子區(qū)域，是否考慮鏈得信息關(guān)系不太大）bedtools getfasta -name -s -fi GRCh38.fa -bed GRCh38.exon.bed >GRCh38.exon.fa# 查看序列信息head GRCh38.exon.fa | cut -c 1-60>ENST00000608838::chr20:87249-87359(+)ACAGGAATTCATATCGGGGTGATCACTCAGAAGAAAAGGTGAATACCGGATGTTGTAAGC>ENST00000608838::chr20:96004-97094(+)GTAGCTTTGAACCCCAAAAATGTTGGAAGAATAATGTAGGACATTGCAGAAGACGATGTT>ENST00000382410::chr20:87709-87767(+)ATGAATATCCTGATGCTGACCTTCATTATCTGTGGGTTGCTAACTCGGGTGACCAAAG>ENST00000382410::chr20:96004-96533(+)GTAGCTTTGAACCCCAAAAATGTTGGAAGAATAATGTAGGACATTGCAGAAGACGATGTT提取一個(gè)個(gè)內(nèi)含子序列
確定內(nèi)含子區(qū)域
sed 's/"/\t/g' GRCh38.gtf | awk 'BEGIN{OFS=FS="\t";oldtr="";}{if($3=="exon") {tr=$14; if(oldtr!=tr) {start=$5; oldtr=tr;} else {print $1,start,$4-1,tr,$20,$7; start=$5;} } }' >GRCh38.intron.bed# 查看文件內(nèi)容head GRCh38.intron.bedchr20 87359 96004 ENST00000608838 DEFB125 +chr20 87767 96004 ENST00000382410 DEFB125 +chr20 142686 145414 ENST00000382398 DEFB126 +chr20 142686 145414 ENST00000542572 DEFB126 +chr20 145488 145578 ENST00000542572 DEFB126 +chr20 157593 158773 ENST00000382388 DEFB127 +chr20 189681 187852 ENST00000334391 DEFB128 -chr20 227346 229277 ENST00000246105 DEFB129 +
提取序列同上。
