Single cell transcriptome workshop

27-28 June 2018

Rémy Bruggmann

Dr. Gharib Walid, UNIBE & SIB

Dr. Tim Tickle, Broad Institute, Cambrigde, USA (Senior product manager)

Mr. Brian Haas, Broad Institute, Cambridge, USA (Senior computational biologist)

Through a combination of lecture materials and hands-on computational exercises, we explore the use of Trinity for de novo rna-seq assembly and downstream analysis of transcript expression, annotation, differential expression, and interactive data analysis. Although participants will be provided with sample RNA-Seq data sets, they are encouraged to bring their own data for use during the guided studies. We can also include information on applications for cancer transcriptome studies and/or studies of non-model organisms lacking reference genomes.

Single cell transcriptome studies are transforming our knowledge about cell types and cell states, and revealing important variation in gene expression that is otherwise hidden in the context of bulk measurements. Single cell RNA-Seq technology and analysis tools are rapidly evolving, and the complexity of such studies necessitates careful statistical considerations. In this workshop, we provide an overview of the sequencing technologies and experimental methods that make possible single cell transcriptome sequencing. Through hands-on activities with single cell RNA-Seq data, we explore analysis methods available for exploring the variation in transcript expression among cells, define clusters of related cells, and identify characteristics of cell types that are relevant to their biological function.

Learning objectives

The overall theme of this course will center on how single-cell RNA-Seq is different than population based RNASeq and, due to those differences, how analysis methodology differs. Participants will have an understanding of how sequence data is generated using the most common single-cell RNA-Seq assays. Participants will form an intuition on how single-cell RNA-Seq expression data is different and how that affects the selection of methodology for analysis. Participants will use popular R libraries to perform quality control, plotting, and analysis targeting both cell heterogeneity and genes that may discriminate those groups.

Schedule overviewDay1: (beginning at 9:15)Morning: Overview of laboratory prep and sequence analysis
Afternoon: Characteristics of expression data and QC

Day2: (beginning at 9:15)Morning: Plotting Single Cell RNA-Seq data
Afternoon: Evaluating and defining cell populations

-----

Day 1: Morning (9:15 am - 12:30 noon)

09:15-10:00 Overview of different wet side preps (SmartSeq2, DropSeq, 10X)

10:00-10:30 Overview of the types of sequences generated from SmartSeq and pipeline for analysis.

10:30-10:40 Break (10 minutes)

10:40-11:30 Overview of DropSeq sequence and analysis pipeline.

11:30-12:00 Overview of 10X sequences and analysis pipeline.

12:00-12:30 Sequence level quality control. Afternoon (1:30 pm - 5 pm)

01:30-02:00 What does single cell expression data look like and why?

02:00-02:30 Introduction to RStudio

02:30-03:00 Initial data exporation

03:00-03:10 Break (10 minutes)

03:10-04:00 Quality control for expression matrices Filtering genes and samples Considerations in data analysis when using UMIs

04:00-05:00 Why normalize gene expression and common types of normalization Using Scone for normalization.

Day 2- Morning (9:15 am - 12:30 noon)

09:15-09:30 Using Seurat to plot genes Plotting (a priori known) marker gene lists to confirm known cell types

09:30-10:30 Why do we need dimensionality reduction and how is this used to plot samples (PCA and tSNE)?

10:30-10:40 Break (10 minutes)

10:40-11:00 Plotting Samples in Seurat.

11:00-12:30 Batch Effects What is a technical batch effect and how to identify them? What new biological batches exist in single cell data? Confounding by study design.

Afternoon (1:30 pm - 5 pm)

01:30-02:30 Moving from clusters to populations of cells (defining clusters given ordinations). Seurat (and RaceID)

02:30-03:30 Differential Expression (SCDE) The different between differential and discriminant expression.

03:30-03:40 Break (10 minutes)

03:40-04:45 Pathway Analysis Pagoda FastProject

04:45-05:00 Overview of available methodology (list what we did not cover). Monocle BiSNE ...etc Resources online for further growth (online tutorials).

UNIBE, Hochschulstrasse 4, room Nr. 028 / EG West

Bring your own data (optional attendance):It is a great opportunity for the participants of the "Single Cell sequencing" and "De novo Transcriptome assembly" workshops to bring their own data and start their analysis onsite and/or ask questions concerning their own research expermiental design.

Prerequisites
Knowledge / competencies:
Attendees should already be familiar with the basic terms and concepts of genetics and genomics.

Technical:
Attendees should bring their own laptop computers.
For the practical tutorial, basic familiarity with the command line environment is required. Basic knowledge in R-statistics is recommended but not mandatory.

Prior to the course, please download the Integrative Genome Viewer (IGV).

This course is free to CUSO PhD students. Participants who are not CUSO PhD students will be charged CHF 120.-.

You will be informed by email of your registration confirmation.

30

Dr. Walid Gharib, Training Group – Swiss Institute of Bioinformatics & IBU - Interfaculty Bioinformatics Unit - University of Bern