本文的分析套路劍指10-20分的journal，全是干貨。
發現一個不錯的專題，講得很詳細。
其中一個細節IDR，講得很通俗。
ATAC-seq data analysis: from FASTQ to peaks
基本概念：
- ATAC-seq和ChIP-seq鑒定出來的peak到底是一些什么區域？除了promotoer就是enhancer
- TSS enrichment和motif是兩個不同的套路，TSS是特指啟動區域，每個基因只有一個；另一個就是motif，這就是抓出TF，哪些TF在發揮作用
分析套路
鑒定差異peak
- 兩組數據比較
- 多組數據比較 - peak cluster
處理細節：
-
Peak calling was performed using MACS2 from all sample reads. 多樣本這樣處理更加靠譜，macs2出來的peak結果才是靠譜的，不要隨便用bedtools修改源頭的peak，否則得到的peak很奇怪
-
The number of raw reads mapped to each peak at each condition was quantified using the intersectBed function in BedTools. 最終用的別人推薦的multicov
-
Raw counts in peaks were normalized using the DESeq package in R.
-
Peak intensity was defined as the log2 of the normalized counts.
bedtools multicov -bams 1.bam 2.bam 3.bam 4.bam-bed file.bed > read.count.txt
可視化差異peak
- peak signal heatmap
- distance to TSS
ngsplot這個工具比較好用，很直觀，有bam就可以用，不像其他工具，一大堆復雜的衍生格式，沒有搞的欲望。
ngs.plot.r -G hg19 -R tss -C example.bam/hesc.H3k4me3.1M.bam -O hesc.H3k4me3.tss -T H3K4me3 -L 3000 -FL 300
- 已經測試通了，用ngsplot來快速畫圖，生成數據；然后用EnrichedHeatmap來畫最終的圖。
- 改一下color set，能夠畫出媲美下面的美圖了。
peak功能注釋
peak鄰近基因表達分析
-
expression level of genes closest to the top 1000 peaks
- 可以畫二維散點圖來展示ATAC-seq和RNA-seq的關系，理想情況是顯著正相關關系。
motif富集分析
- TF motifs enriched in peak clusters
- 結合轉錄組基因表達數據驗證
- 如何直接在ggplot里添加motif images，教程
- 可以直接用meme-chip一步到位
export PATH=/home/lizhixin/softwares/ATAC-seq-conda/anaconda3/bin:$PATH sortBed -g genome.txt -i up.sig.peak.bed > sorted.up.sig.peak.bed sortBed -g genome.txt -i down.sig.peak.bed > sorted.down.sig.peak.bed cut -f1-3 sorted.up.sig.peak.bed > sorted.up.sig.peak.3col.bed cut -f1-3 sorted.down.sig.peak.bed > sorted.down.sig.peak.3col.bed bedtools getfasta -fi /home/lizhixin/softwares/miniconda3/share/homer-4.10-0/.//data/genomes/hg38///genome.fa -bed sorted.up.sig.peak.3col.bed -fo sorted.up.sig.peak.3col.fasta bedtools getfasta -fi /home/lizhixin/softwares/miniconda3/share/homer-4.10-0/.//data/genomes/hg38///genome.fa -bed sorted.down.sig.peak.3col.bed -fo sorted.down.sig.peak.3col.fasta intersectBed -a /home/lizhixin/project2/analysis/ATAC-seq/encode-pipeline/ipsc/result/atac/c3097359-919e-43d3-9b81-551e4c8ff029/call-filter/shard-0/execution/glob-3bcbe4e7489c90f75e0523ac6f3a9385/IMR90-iPS_1.trim.merged.nodup.no_chrM_MT.bam -b sorted.up.sig.peak.3col.bed > tmp.bam samtools index tmp.bam ngs.plot.r -G hg19 -R tss -C tmp.bam -O IMR90-iPS_1.tss -T IMR90-iPS_1 -L 3000 -FL 300
motif印跡分析
- ATAC-seq footprint for specific motifs
- rgt-hint這個工具可以做，教程
- 搞明白什么是 TF footprint，看這篇文章：Identification of transcription factor binding sites using ATAC-seq
轉錄因子網絡
- Cis-regulatory networks of TFs
ChIP-seq數據驗證ATAC-seq結果
個別基因可視化問題 - bigwig
- UCSC genome browser看signal，都是導出pdf，自己再編輯
- IGV
- 測序數據可視化 （三） - UCSC genome browser
重點文獻：
- From reads to insight: a hitchhiker’s guide to ATAC-seq data analysis
- Chromatin accessibility analysis reveals regulatory dynamics of developing human retina and hiPSC-derived retinal organoids