Complex and Dynamic Analysis of Nascent Transcription Introduction

Complex and Dynamic Analysis of Nascent Transcription Introduction and overview

Robin Dowell Bio. Frontiers Institute University of Colorado Boulder Robin. Dowell@colorado. edu Margaret. Gruca@colorado. edu Jacob. Stanley@colorado. edu Margaret Gruca Nascent analysis Jacob Stanley RNA-seq

Goals of this workshop 1. Understand the distinction between nascent transcription and steady state RNA (RNA-seq) analysis. 2. Work through a comparative transcriptome data analysis example. 3. Discuss basic statistical concepts that are paramount to understanding your data.

Schedule: Today • Session 1 (1: 30 pm - 2: 45 pm): Studying Transcription -- an Overview • Coffee Break (2: 45 -3: 15 pm) • Session 2 (3: 15 -4: 30 pm): Garbage In, Garbage Out: Quality control on sequence data • Session 3 (4: 30 -5: 15 pm): Maximize read usage through mapping strategies

Schedule: Tuesday Morning • Session 4 (8: 30 -9: 30 am): Learning to count: quantifying signal • Coffee Break (9: 30 -10: 00) • Session 5 (10 -11: 15) : Assessing changes in data (part I: resorting to statistics) • Session 6 (11: 15 -12: 30): Assessing changes in data (part II: differential expression)

Schedule: Tuesday Afternoon • Session 7 (1: 30 -2: 45): Annotation agnostic analysis: finding new things in the data • Coffee Break (2: 45 -3: 00 pm) • Session 8 (3: 15 -4: 30): Getting the most from your data: Integration with relevant other data

Key concepts of workshop • Understanding the experiment is key to its analysis. • A fundamental principle in data science is understanding your expectation and then evaluating your reality. • Reproducibility requires keeping good records. You *MUST* report on your software choices, options, order, etc.

Our generic differential transcription pipeline

Every step has important choices to be made.

There is no ONE RIGHT answer. • Every algorithm/pipeline is different • Different assumptions • Order of processing can sometimes have a significant effect • Different options can dramatically change the result • MOST projects require customization and personalized attention to get useful data

Important consideration for bioinformatics (in general) Keep a digital notebook • • • Program and version Options and variables utilized Order of usage Specifics of compute environment Avoid “manual edits” whenever possible Backup your work!! You must be able to report on these elements of your analysis in order for your work to be REPRODUCIBLE!!!!

https: //rmghc. slack. com/

Our compute environment Amazon Web Services (https: //aws. amazon. com) graciously provided resources for the workshops and hackathon. We are using an EC 2 compute instance setup by Bio. Frontiers IT staff. https: //docs. google. com/spreadsheets/d/14 v. Stf 4 em. JO 6 i. REK 8 Vsy. SLa 7 O 45 vb 8 y. Rf. LOFm 3 NQIZuo/edit#gid=0

Assumptions of this workshop 1) You start with a bag of reads off the sequencer, BUT … MANY decisions pre-sequencing influence your analysis. So we MUST understand how the sample was handled BEFORE it was sequenced. 2) You already know how to use a Unix/Linux command line

http: //dowell. colorado. edu/Hack. Con/pages/hackcon. html

To view on your laptop … X 2 Go http: //dowell. colorado. edu/Hack. Con/pages/hackcon. html

What is IGV A desktop/server application for integrated visualization of multiple data types and annotations in the context of the genome Microarrays Epigenomics RNA-Seq NGS alignments Comparative genomics

IGV layout Cytoband Track Names Genomic Coordinates Data Panel Annotation Heatmap Genome Features

Every cell contains essentially the same DNA, but is distinct

But how? Transcriptional Regulation

Despite years of expression studies, we have NOT been assaying transcription… Nascent Transcription Steady state RNA = Expression studies

Transcriptional Regulation Gene locus (DNA) transcription pre-m. RNA Nascent transcription productive splicing productive m. RNA translation Protein RNA-seq

Transcription cycle driven by RNA polymerases Image: Adapted from Fuda et. al (2009)

Nascent Transcription: isolating and sequencing nascent RNA nuclei Global Nuclear Run On Sequencing (GRO-seq) +ATP, GTP, CTP, Br. UTP = Grow cells Polymerase paused Nuclear run-on (5 -10 min) α Sequence 50 nt single end 150 -300 nt fragments Library prep α α α Isolation 2 x Base hydrolysis Visualize & Analyze Map reads to genome Core and Lis (2008)

Three Classes of Transcription in Eukaryotes RNA polymerase I (pol I) ribosomal RNAs (5. 8 S, 18 S, 28 S r. RNA) RNA polymerase II (pol II) m. RNAs some small nuclear RNAs (sn. RNAs) non-coding RNAs (mostly of unknown function) RNA polymerase III (pol III) t. RNAs 5 S RNA some sn. RNAs small cytoplasmic RNAs (sc. RNAs) Nascent transcription tag the newly synthesized RNA – hence they obtain reads from ALL THREE classes of polymerase.

Splicing is co-transcriptional RNA Pol II

Sequence signals orchestrate the process End of transcription? Nascent transcription (PRO-seq, GRO-seq) RNA-seq (numerous variations)

m. RNA-seq: isolating and sequencing mature m. RNA Convert to ds-c. DNA and shearing poly-A selection or ribosomal subtraction Amplification and adapter ligation

A data example

So what kinds of problems are most appropriate to each protocol?

But regardless of which transcription protocol (nascent or steady state), there are important consistent considerations! This is the single largest goal of this workshop!!

Schedule: Today • Session 1 (1: 30 pm - 2: 45 pm): Studying Transcription -- an Overview • Coffee Break (2: 45 -3: 15 pm) • Session 2 (3: 15 -4: 30 pm): Garbage In, Garbage Out: Quality control on sequence data • Session 3 (4: 30 -5: 15 pm): Maximize read usage through mapping strategies