Shaun Jackman

Genome Sciences Centre, BC Cancer Agency
Vancouver, Canada
@sjackman
github.com/sjackman
sjackman.ca

Thesis committee

Inanc Birol
Joerg Bohlmann
Jenny Bryan
Steven Hallam

Chair

Steven Jones

Conferences

rOpenSci Unconference
RECOMB 2015

Teaching assistant

STAT 545 Data wrangling, exploration, and analysis with R
Automating data analysis pipelines

BIOF 520 Problem-Based Learning In Bioinformatics
Genomic epidemiology

• UniqTag: Content-derived unique and stable identifiers for gene annotation
  SD Jackman, J Bohlmann, I Birol
  PLOS ONE 2015
• Sealer: a scalable gap-closing application for finishing draft genomes
  D Paulino, RL Warren, BP Vandervalk, A Raymond, SD Jackman, I Birol
  BMC Bioinformatics 2015
• Improved white spruce (Picea glauca) genome assemblies and annotation of large gene families of conifer terpenoid and phenolic defense metabolism
  RL Warren, CI Keeling, MMS Yuen, A Raymond, GA Taylor, …
  The Plant Journal 2015
• Spaced Seed Data Structures for De Novo Assembly
  I Birol, J Chu, H Mohamadi, SD Jackman, K Raghavan, BP Vandervalk, …
  International Journal of Genomics 2015
• DIDA: Distributed Indexing Dispatched Alignment
  H Mohamadi, BP Vandervalk, A Raymond, SD Jackman, J Chu, …
  PLOS ONE 2015
• On the Representation of De Bruijn Graphs
  R Chikhi, A Limasset, SD Jackman, JT Simpson, P Medvedev
  Journal of Computational Biology 2015
• BioBloom tools: fast, accurate and memory-efficient host species sequence screening using bloom filters
  J Chu, S Sadeghi, A Raymond, SD Jackman, KM Nip, R Mar, …
  Bioinformatics 2014
• Konnector: Connecting paired-end reads using a bloom filter de Bruijn graph
  BP Vandervalk, SD Jackman, A Raymond, H Mohamadi, C Yang, D Attali, …
  Bioinformatics and Biomedicine (BIBM) 2014

White Spruce Organelles

Organellar Genomes of White Spruce (Picea glauca): Assembly and Annotation

ABySS

Assembly by Spaced Seeds

for the assembly of long reads

DistanceEst

Estimating the distance between two contigs

Homebrew Science

Homebrew | Linuxbrew | Homebrew-science

Thesis proposal

1. Organellar Genomes of White Spruce (Picea glauca): Assembly and Annotation
2. Scale ABySS to assemble long, accurate reads
3. Scaffold using SMS and molecular barcoding
4. Assemble single-molecule sequencing data

Organellar genomes

Completed

• Assembled cpDNA and mtDNA genomes
• Annotated genes (mRNA, rRNA, tRNA) and repeats
• Analysed RNA-seq data to quantify
  • transcript abundance in eight tissues
  • expressed ORFs
  • C-to-U RNA editing

Organellar genomes

Plan

• Analyse RNA-seq data to annotate
  • cryptic ACG start codons
    due to C-to-U RNA editing
  • spliced ORFs
• Submit annotated genomes to GenBank
• Complete manuscript

Scaling ABySS

For the assembly of long reads

• 500 bp overlapping paired-end reads (MiSeq)
• 10 kbp synthetic long reads (Moleculo)
• 8–200 kbp corrected single-molecule sequencing

Using memory efficient data structures

• spaced seed de Bruijn Graph
• Bloom filter de Bruijn Graph

Scaling ABySS

Completed

• Implemented Bloom filter de Bruijn Graph
  • Konnector with Ben
  • Sealer with Daniel
• Implemented spaced seed de Bruijn Graph with Karthika
• Reduced memory and time requirements for long reads
• Memory usage is independent of parameter k

Plan

• Assemble model organisms, human in particular, with ABySS 2.0
• Compare to ABySS 1.5 and other dBG assemblers

RECOMB 2015

Research in Computational Molecular Biology
Warsaw, Poland · 2015 April 10–15

Scaffolding

Order and orient contigs to build scaffolds using…

Completed

• Illumina mate-pair libraries
• Illumina Moleculo with Tony

Plan

• Single-molecule sequencing
• Molecular barcoding
  10X Genomics GemCode
  for white spruce (Picea glauca)

Assemble SMS

Drosophila melanogaster

• Assembly with PBcR used 621,000 CPU-hours,
  or 26 days with 1000 cores
• MHAP reduced assembly time to 700 CPU-hours
• Efficient algorithms make a big difference!

Homo sapiens

• Assembly of CHM1 with HGAP used 405,000 CPU-hours
  in one day on Google Cloud!
• More work to be done for PacBio
• And lots more work to be done for Nanopore

Assemble SMS

Overlap

• Spaced seed clustering with Hamid and Justin

Layout

• Numerical solution with ABySS-DrawGraph

Consensus

• BWA-MEM and samtools
• Assembled a small 100 kbp synthetic genome

Plan

• Continue development with Hamid, Justin and Ben
• Assemble Nanopore Escherichia coli (5 Mbp)
  • Nanopore Saccharomyces cerevisiae (12 Mbp)
  • PacBio Homo sapiens CHM1 (3 Gbp)

Thesis proposal

1. Organellar Genomes of White Spruce (Picea glauca): Assembly and Annotation
2. Scale ABySS to assemble long, accurate reads
3. Scaffold using SMS and molecular barcoding
4. Assemble single-molecule sequencing data

Shaun Jackman

Genome Sciences Centre, BC Cancer Agency
Vancouver, Canada
@sjackman
github.com/sjackman
sjackman.ca

Links

Inanc Birol | Joerg Bohlmann
Steven Hallam | Jenny Bryan
RECOMB 2015 | rOpenSci Unconference
UniqTag
White spruce organelles | ABySS 2.0 | DistanceEst
BIOF 520 | STAT 545
Open, reproducible science
Homebrew | Linuxbrew | Homebrew-science

Software

BLASR | BWA-MEM | Celera Assembler
DALIGNER | Dazzler | Falcon | HGAP
LAST | MHAP | Nanocorr | Nanocorrect
Nanopolish | PBcR | PBDAGCon | POA
Quiver

PacBio RS II

Oxford Nanopore MinION

Sequencing technologies

A visual comparison

PacBio circular consensus

Nanopore 2D

Assembly Stages

Overlap

Find all significantly overlapping reads

Correct

Recall the consensus base of each read

Layout

Determine the order and orientation of the reads

Consensus

Call the consensus base of each contig

Tools

Overlap

BLASR · DALIGNER · MHAP

Correct

PBDAGCon · Falcon · Dazzler · Nanocorrect

Layout

Celera Assembler · Falcon · Dazzler

Consensus

Quiver · Nanopolish

Pipelines

Assembling large genomes

Assembling small genomes in 2015

Nanopore + Illumina

Nanocorr Saccharomyces cerevisiae (12 Mbp)

Nanopore only

Nanopolish Escherichia coli (5 Mbp)

Retool for SMS

Alignment

Seed and extend is difficult when
one in seven bases is incorrect

Assembly

de Bruijn graphs also require accurate seeds
Return to overlap, layout, consensus (OLC)

Align

• BLASR PacBio
• BWA-MEM (Burrows-Wheeler Aligner) Heng Li
• LAST Computational Biology Research Consortium, Tokyo, Japan

Overlap

• BLASR PacBio
• DALIGNER Gene Myers
• MHAP (MinHash Alignment Process) Adam Phillippy

Read correction

• PBDAGCon (PacBio DAG Consensus) PacBio
• Falcon PacBio
• Dazzler (Dresden Azzembler) Gene Myers
• Nanocorrect Jared Simpson (using POA)

Layout

• Celera Assembler J. Craig Venter Institute (JCVI)
• Falcon PacBio
• Dazzler (Dresden Azzembler) Gene Myers

Consensus

• Quiver PacBio
• Nanopolish Jared Simpson

Assembly Pipeline

• HGAP (Hierarchical Genome Assembly Process) PacBio
• Falcon PacBio
• PBcR (PacBio Corrected Reads) Celera Assembler, JCVI
• Dazzler (Dresden Azzembler) Gene Myers
• Nanocorr Mike Schatz
• Nanopolish Jared Simpson