This repo contains all the material for the computer labs associated with the two courses CB2040 or BB2255 (which overlap to a large extent); the only difference between the two is that the latter (BB2255) has one less lab session, meaning that:
- Students from CB2040 should do : Lab 1, Lab 2, Lab 3 and Lab 4
- Students from BB2255 should do : Lab 1, Lab 3 and Lab 4
You will find the all the essential information regarding these labs within this repo, such as guidelines regarding the execution of the labs, contact information, deadlines and brief descriptions of the labs with links to material that might be of help to you.
The slides from the first lecture (introduction) can be found here.
-
What is R?: R is an interpreted programming language, just like python, in short meaning you do not have to compile your code in order to run it. R was first released in 1993, it has become widely popular among statisticians due to a vast array of libraries which allows for seamless implementation of several statistical techniques such as GLM's (generalized linear models), classical statistical testing, clustering and classifiers. Bioinformatics in large can be considered as statistics applied to biological data, hence R with it's rich library of tools for statistical analysis has become popular to use among bioinformaticians.
-
Why use R?: The usefulness of a language is to a large extent dependent on the community, and people contributing with libraries. Whilst R as an language in no way is superior to for example Python or Julia, as of now the set of available tools for bioinformatic analysis in R by far exceeds the others. This allows you to conduct advanced analyses with just a few lines of code rather than several hundreds or thousands. Popular machine learning frameworks such as TensorFlow have also started to provide an interface for usage in R, meaning that more complex statistical inference using techniques from machine learning like CNN's, Autoencoders and LSTM's now can be utilized with ease as well.
-
Where to get R?: You can download R to your personal computer, following the instructions at the official webpage. If you are unfamiliar with R or do not have an extensive programming background using the IDE Rstudio is something we recommend - though you are of course free to use your favorite IDE or text manager (like Vim or Emacs).
-
Where to start? The first computer lab will focus on learning the basics of R, and we will assume that you have zero previous experience with the language, hence that could figure as your first starting point. Nevertheless, if you want to place yourself in a very good position and gain somewhat of a headstart we'd recommend you to have a look at some of the online tutorials out there, there's plenty to choose from ("R programming tutorial" has
1 390 000 000hits on google... ), so at least one should be compatible with your preferred ways of learning.
Main Responsible TA: Alma Andersson
Lab Status: Ready
Here you will familiarize yourself with some of the basic syntax and documentation in R. This lab will lay the foundation for the two next labs, where the focus will be more on analyzing data with the help of certain bioinformatic packages. Thus it is imperative that you put effort into learning how to orient yourself in R.
Before Lab 1 starts make sure that you:
- Install R
- Clone this repo (see below)
- Install R packages for Lab 1 (see below)
- Install Rsdtudio or an IDE that can process R Markdown files
execute the above steps in the order that they are given.
Git is a system developed for distributed version control, which basically means that it prevents people working on the same project from overwriting each other's code, you can think of it as a sort of "Google Docs" for programming (but way more sophisticated).
You will create a copy the code that we (the TAs) have put up here by cloning this repository. To do this, open the terminal and:
- Orient yourself to a directory where you want to put all the material
- Run the command
git clone https://github.com/almaan/genetech.git
The files will be downloaded and found in a subfolder of your current directory
called genetech.
As mentioned above, the real value of R comes with its large ecosystem of packages which contain several nifty features that makes it easier to conduct our analysis. In short, a package is code written by someone else that spares you from having to write thousands of lines of code.
While packages saves us a lot of work, they can also cause some headaches - mainly when one is trying to install them. To - hopefully - make this process slightly less painful for you, we've prepared an installation script for you. Thus what you should do in order to install all the packages you need for this course, is to:
- Install R (you cannot install R packages before you have installed R)
- From the terminal run the file
install-packages.R, to do so:- Open the terminal
- Go to the directory of the cloned repo (the one called
genetech) - Change directory to
labs/prep - Run the command
Rscript --vanilla .install-packages.R
In addition to changing the author information (see Guidelines section below),
also change the variable GRADE_MODE from FALSE to TRUE before handing in
the lab, if your lab knitts without any complications, you have solved all the
exercises correctly.
Main Responsible TA: Sami Saarenpää
Lab Status: Ready
This lab is divided into two parts.
In the first part of the lab we will explore RNA-seq data analysis. After the first publication in 2009 (Wang et.al) this method has allowed researchers to study gene expression in RNA level introducing a baseline for methods developed after like single cell RNA-seq and spatial transcriptomics.
In this part we will identify differentially expressed genes of different kinase inhibitors in treated mice tumors. We will use DESeq2 for differential gene expression analysis, visualise the results with heatmap and identify some of these genes.
In the second part we will explore variant calling and GWAS which are commonly used to analyse different traits on genomic level.
Variant calling identifies single nucleotide polymorphisms (SNPs) and small insertions and deletions (indels) from NGS data. There are available plenty of tools to search for variants from genomes and they have become a standard in population based studies. In this lab we will analyse variants using genome wide association study (GWAS). In GWAS the genetic variants in different individuals are analysed to see if any variant is associated with a trait like major diseases in humans or flowering time in plants. Basically, GWAS tries to find correlation between phenotype and changes in genomic areas on a population level.
Because this part takes a long time to compute, the output of the code is already available. Instead the focus is to understand what type of data one could get out of variant analysis and GWAS analysis. If you want later to use the code, you will be able to run it.
You will hand in one knitted html file in Canvas latest one week after the lab. Remember to change the author information before handing in the lab.
Main Responsible TA: Ludvig Larsson
Lab Status: Ready
In this lab you will be working on two different single-cell datasets. Single-cell RNA sequencing (or scRNA-seq) have exploded in the last decade and has become one of the most valuable tools in genomics to define celltypes, explore cell states, study cell to cell interactions, track differentiation processes over time, study response to drugs and much more. ScRNA-seq methods have also become important tools in ongoing research efforts to create atlases of all the organs in the human body and we are still learning more about new cell types and cell states every year. In this lab we will use the popular Seurat R package to explore some of the most essential parts of a scRNA-seq analysis workflow.
Main Responsible TA: Alma Andersson
Lab Status: Ready
In this lab you will work with Visium data, looking at parts of the mouse brain! While similar to single cell data in some ways, there are also some important differences. One thing that's extremely attractive with ST-data is that the very design of the method allows us to visualize the gene expression in the physical 2D plane - which is one of the exercises in this lab. You will apply some of the concepts introduced in the previous labs, in order to make sense of your spatial data, but also explore new ones like LDA and Single Cell Data Integration.
We will use quite a lot of different packages in this lab, some which may take a
while to install. Thus, it might be a good idea to install these before the lab
starts, as to make sure that you don't waste valuable time waiting for something
to compile. Open the script prep/install-packages.R in Rstudio and run the
whole thing, just to make sure you have all the necessary components installed.
You should more or less have all tools required to conduct this lab on your computer, meaning you won't have to install plenty of new packages. However, we recommended that you read up a bit on the following concepts:
- Spatial Transcriptomics (The method presented here)
- Dimensionality Reduction (Focus on PCA)
- Clustering
- LDA (Latent Dirichlet Allocation)
- Single Cell Integration : link
All labs should be handed in via Canvas, they should be provided as a
knitted html file, raw markdown files will not be corrected but rather
sent back for you to knit.
All labs will start with something similar to this:
---
title: "Lab 1 - Introduction to R"
author: "Alma Andersson"
date: "11-09-2020"
output:
tufte::tufte_html: default
/---
When you hand in the modified lab, change the author field to your name(s).
Programming and this type of analysis is best learnt by doing; hence we encourage and strongly recommend that you work alone. But you may work in pairs if that is what you prefer, however groups of more than two people are not accepted. If you do work in pairs, please state who you are working with in the report and comment section of your hand in, also make sure BOTH of you hand in a copy of the report (identical).
-
Your Deadlines: You will have one week to complete each lab and hand in the report for that specific lab. Meaning if your lab is on Monday week T then it should be handed in no later than Monday 23:59:59 week T+1. Red days and holidays (aside from Saturday and Sunday) will be taken into consideration.
-
Our Deadlines: We will make sure to grade all reports within one week after your deadline (with exception for late reports, see below). Meaning if your lab is on Monday week T then it should have been corrected no later than Monday 23:59:59 week T+2. Red days and holidays (aside from Saturday and Sunday) will be taken into consideration.
-
If you hand in your report late, we will still correct it. But you will be given an extra assignment to complete, information regarding the deadline for this assignment is given by us if such a scenario arise. Finally, note that a report is only considered as handed in when all questions have been answered - meaning that handing in a report with only 50% of the questions answered before the deadline will not be considered as "in time".
-
If you hand in a report that we do not consider of high enough quality, (i.e. questions have been answered incorrectly), you will be given a second attempt to complete the report. We will accept two iterations of this procedure (i.e. you will be given two chances to hand in a sufficiently good report additional to your first). If you after two attempts still have not provided us with a report of acceptable quality, you will be given an extra assignment, along with more explicit details of what you need to change in order for your original report to pass.
- Alma Andersson : almaan [at] kth.se
- Sami Saarenpää : sami.saarenpaa [at] scilifelab.se
- Ludvig Larsson : ludvig.larsson [at ] scilifelab.se