Bulk RNA-seq Data Analysis using High-Performance Computing (bulk RNA-seq Part I – FASTQ to counts)

Learning Objectives

• Understand the necessity for, and use of, the command line interface (bash) and HPC for analyzing high-throughput sequencing data.
• Understand best practices for designing an RNA-seq experiment and analysis the resulting data.

Installations

All:

• FileZilla Client (make sure you get ‘FileZilla Client’)

Mac users:

• Plain text editor like Sublime text or similar

Windows users:

• GitBash
• Plain text editor like Notepad++ or similar

Notes

• These materials focus on the use of local computational resources at Harvard, which are only accessible to Harvard affiliates
• Non-Harvard folks can download the data and set up to work on their local clusters (with the help of local system administrators)

Instructions for Harvard researchers with access to HMS-RC’s O2 cluster

To run through the code in the lessons below, you will need to be logged into O2 and working on a compute node (i.e. your command prompt should have the word compute in it).

1. Log in using ssh ecommonsID@o2.hms.harvard.edu and enter your password.
2. Once you are on the login node, use srun --pty -p interactive -t 0-2:30 --mem 1G /bin/bash to get on a compute node or as specified in the lesson.
3. Proceed only once your command prompt has the word compute in it.
4. If you log out between lessons (using the exit command twice), please follow points 1. and 2. above to log back in and get on a compute node when you restart with the self learning.

Lessons

Part 1

1. Introduction to RNA-seq
2. Shell basics review
3. Working in an HPC environment - Review
4. Best Practices in Research Data Management (RDM)
5. Project Organization (using Data Management best practices)

---

Part II

1. Quality Control of Sequence Data: Running FASTQC
2. Experimental design considerations
3. Quality Control of Sequence Data: Running FASTQC on multiple samples
4. Quality Control of Sequence Data: Evaluating FASTQC reports

---

Part III

1. Sequence Alignment Theory
2. Quantifying expression using alignment-free methods (Salmon on multiple samples)

---

Part IV

1. QC with Alignment Data
2. Documenting Steps in the Workflow with MultiQC
3. Troubleshooting RNA-seq Data Analysis

---

Part V

1. Automating the RNA-seq workflow

---

Answer Keys

• Experimental design (one possible solution)
• FASTQC sbatch script
• FASTQC sbatch script .out file
• FASTQC sbatch script .err file.
• sbatch script to run salmon for all samples
• Automation Script

---

Building on this workshop

• Introduction to R workshop materials
• Bulk RNA-seq Part II (differential gene expression analysis) materials

---

Resources

• Video about statistics behind salmon quantification
• Advanced bash for working on O2:
  • Creating shortcuts or aliases
  • Copying files from other remote locations to O2
  • Creating symbolic links
• Obtaining reference genomes or transcriptomes

---

These materials have been developed by members of the teaching team at the Harvard Chan Bioinformatics Core (HBC). These are open access materials distributed under the terms of the Creative Commons Attribution license (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.