RNA-seq workshop-Day 1

Day 1.jpg

本周的Data Workshop又开始了，这次将围绕着以R语言为工具，进行RNA-seq和ScRNA-seq的分析。今天主要回顾了R introduction的内容，温习了接下来将要用到的一些commands，然后对RNA-seq的流程进行了系列介绍。

1. Introduction to R (Dr. Rocio T Martinez-Nunez）

1.1 Objects

• Assign to objects(vectors, tables, values, functions)

1.2 Commenting your code

• Just add (#) before what you want to comment

1.3 system(): communicates with the shell in your computer

system("ls -F/")

1.4 cmd as a group of commands

cmd <- paste("gunzip -c", fastq.files, "| head")
cmd  # to view cmds & runs
system(cmd[1]) # Run the first command of cmd

1.5 Some R tips

1.5.1 ask for help

# in R: ? + function
?system
#in shell : (-h)
system("trim_galore -h")

1.5.2 Tab: look for the list of word match in R.

1.5.3 Arrow keys: up row-the last thing you type in.

1.5.4 Pines %>% in R or | in shell

install. packages("tidyverse")  # install packages
library("tidyverse")  # load packages
download.file("website", "path and name. csv")  # download file
surveys <- read_csv("path and name. csv")  # open file
str( surveys)  # inspect the data: an overview of an object's structure and its elements
dim( surveys)  # size: row numbers and column numbers
head( surveys)  # check the top(first six lines) of the data frame
surveys_new <- surveys %>%  # pipes
filter(weight < 5) %>%  # filter
select(species_id, sex, weight)  # select
str(surveys_new)  # inspect the data: an overview of an object's structure and its elements
dim(surveys_new)  # size: row numbers and column numbers
head(surveys_new)  # check the top(first six lines) of the data frame

• Only works when install tidyverse.
• %>% : shortcut keys in PC: ctrl + shift + M
• %>% means then, (the things we want pipe) on the left, and (the things we want to pine into) on the right.

1.6 Some R functions we will be using:

 # create command cmd that includes trim_galore and its flags with the object we apply it to   
cmd <- paste("trim_galore --length 21 --output_dir trimgalore, fastq.files)  
# run only the first line of the commands
system(cmd[1])
# create vector with the power of 1, 2 and 3:
sapply(1:3, function(x) x^2)
#[1] 1, 4, 9

• system(): communicates with the shell.
• dir.create(): create directories.
• list.files(): list the files in your working directory.
• paste(): concatenates vectors after converting into character.
• data.frame(): generates a data frame.
• sapply(): applies a function to an object and returns a simplified object.

1.7 Loops: vectorization & sapply

for (year in c(2010, 2011, 2012, 2013, 2014, 2015)){
      print(paste("The year is", year))
}

2. Introduction to RNA-seq data analysis (Dr. Alessandra Vigilante)

2.1 What is NGS

• Next-generation sequencing (NGS), also known as high-throughput sequencing, is the term used to describe a number of different modern sequencing technologies, such as RNA-seq, ScRNA-seq, ChIP-seq et al.

2.2 Eight stages in RNA-seq Analysis

2.2.1 Define the question of interest (RNA-seq data can tell us)

• Relative expression levels within a biological sample
• Gene expression differences between biological samples
• Quantify alternative transcript levels
• Confirm annotated 5′ and 3′ ends of genes
• Map exon/intron boundaries

2.2.2 Get the data(data formats)

• Raw data: Fastq
• Aligned data: SAM, BAM, CRAM
• Genome annotation: GFF
• Intervals: BED
• Variants: VCF, BCF

2.2.3 Clean the data(quality control)

• FastQC: trimmomatic, cutadapt
• The ShortRead package in R/Bioconductor using the qa() and report () functions

2.2.4 Map the data

• Chanllenges: large costs in memory; introns; updates of reference genomes, tools and softwares.
• Mapping srategies: de novo assembly, align to transcriptome, align to genome.
• Tools: Bowtie 2, TopHat 2, STAR
• Pseudo-alignment: Kallisto - faster and more accurate
• If you have SAM files you have to transform them to BAM
• You can visualise your BAM files in IGV
• Use either your BAM file or the transcript abundance file (from Kallisto) to
  generate a Count Table
• Perform differential expression analysis and downstream analyses

2.2.5 Explore the data

2.2.6 Fit statistical models

2.2.7 Make your analysis reproducible

RNA-seq workflow in the workshop

3. Learning experience

• 今天第一个到workshop，一切准备很充分，全天学习很投入。
• 今天课程比较杂，遇到的很多新的问题和挑战，需要好好消化。
• 今天认识了Guys Campus的口腔医学华人博士，聊得很开心，KCL的口腔医学已经世界排名第二啦，进一步了解了国外博士的生活和学习风貌，值得学习他们的新技术新方法。
• 今天还认识了Denmark Campus的生信大牛，乐于助人还给我们讲述他的学习历程，希望接下来可以继续向他们请教，互帮互助。

本次笔记借鉴了KCL Workshop的学习资料及课件，请勿转载，如需引用请注明。