miRNA-seq数据分析其实大同小异

生信菜鸟团 · 公众号 · 生物 · 2020-10-27 17:20

正文

有粉丝咨询，他感兴趣的领域的一篇文章，提供了miRNA-seq数据，但根据我的GEO数据挖掘课程代码，没办法下载到表达量矩阵文件。如下：

GEO数据挖掘课程代码没办法处理测序数据

希望我出一个miRNA-Seq数据实战视频课程，因为这篇文章的其原始数据上传到了公共数据库，见：https://www.ncbi.nlm.nih.gov/sra?term=SRP111349。恰好可以用来做课程示例。

GEO数据挖掘视频观看方式（附带交流群）

然后 马上就有了3千多学习量 ，而且有学员给出来了图文并茂版本万字笔记，让我非常感动！

扫描下面二维码马上就可以学习起来啦，笔记需要至少半个小时来阅读哦！

使用我的GEO数据挖掘课程代码，我提供的例子是针对表达量芯片的，并不是针对测序数据。但并不是说，测序数据就不能分析，也并不是说它就没有提供表达矩阵。

文章对应的表达矩阵也是公开可以获取的，https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE100932

其实这篇文章（URL: https://doi.org/10.1007/s12217-018-9653-2 ）有3种数据：

关于miRNA-seq数据分析流程 流程，文献描述如下：

Pre-processing: Sequencing adapters removal performed with Trimmomatic (v0.32)
Pre-processing: reads in the size range of 18-50 nucleotides were selected and used for downstream analysis
Mapping: mapped against mature miRNAs in the mirBase (version 21) using BWA tool (command bwa aln; v0.7.5a)
Feature count: miRNAs expression counts were obtained with samtools (z1.2)
Differential expression analysis: differential expression analysis was performed using the edgeR (Bioconductor version 2.13). Counts were normalized and differential expression was inferred using the negative binomial distribution and a shrinkage estimator for the distribution variance of the counts.
Genome_build: mirBase: r21 mature RNA
Supplementary_files_format_and_content: raw count of reads mapping to different miRNAs for each sample

**欢迎共享哦，**比如大家可以看到的tcga数据库的mRNA Analysis Pipeline ，详细代码：

https://docs.gdc.cancer.gov/Data/Bioinformatics_Pipelines/Expression_mRNA_Pipeline/

太复杂的流程就算了，比如上面提到的发表在 Nucleic Acids Res. 2016 Jan 8; 的文章的流程：

普通人一辈子也就是处理两三次miRNA数据，并不是TCGA计划那样专业的团队，所以我们仅仅是关心测序reads的清洗问题，接头去除，以及比对的策略。定量之后的表达矩阵分析，反而是很简单的。欢迎分享你验证过的拿手的流程哈，发邮件给我，到 [email protected]

前面我提到过，这个文章有3种数据，其中对应的RNA-seq数据分析 流程描述 如下：

Pre processing: Sickle trimmer was used to remove low quality ends of the raw reads (using a quality threshold of 20, with option –q 20) and to remove short reads (using a minimum length threshold of 30, with option –l 30); Adapters were removed using CUTADAPT
Mapping: Reads were firstly aligned to reference genome with TopHat2 (v2.1.1); unmapped reads were re-aligned with Bowtie2 (v2.2.9) in local mode. Alignments were joined with picard tools (v1.119)
Feature count: Gene count data were obtained with htsseq-count command in HTSeq framework (v0.6.0)
Differential expression analysis: performed using the edgeR (v3.16.5) package in R. After library normalization, the genewise exact tests for differences in the mean between the two groups of negative-binomially distributed counts was calculated with command exactTest. Significantly differentially expressed genes (DE-genes) between the two conditions (FMOM vs FMSM), without adjustment for multiple testing, were identified and exported for further analysis.
Genome_build: ENSEMBLE GRCh38.p5 Human Genome assembly
Supplementary_files_format_and_content: text file with mRNA count data.

感兴趣可以细读表达芯片的公共数据库挖掘系列推文；

虽然表达矩阵的获取是类似的（不管是表达量芯片，还是测序），但是 下游分析 千差万别：

而miRNA-seq数据虽然也是表达量矩阵，差异分析，差异miRNA列表，但是miRNA本身是没有后续分析的丰富的数据库资源的。之前我在生信技能树分享了几个miRNA的靶向基因的查询工具，分别是：

如果你把你的目标miRNA的靶基因找到了，就可以走差不多的下游分析，包括KEGG,GO等生物学功能数据库的注释。