Using the Iso-Seq Application on SMRTlink and BioConda

Elizabeth Tseng, Principal Scientist, PacBio

Why use Iso-Seq analysis?

ISO-SEQ ANALYSIS MAIN FEATURES

• No reference genome required
• No transcriptome assembly required
• Recovers full-length (5’ to 3’) transcripts
• Yields highly accurate (>99%) transcripts

HIFI READS FROM CCS

FULL-LENGTH READS HAVE 5’ AND 3’ PRIMERS

REMOVE CONCATEMERS AND POLY(A) TAILS

CLUSTER TO GET ISOFORMS

• High Quality (HQ): accuracy ≥99% AND ≥2 FLNC read support
• Low Quality (LQ): accuracy <99% OR <2 FLNC read support

Note: The pdf is wrong. Low quality is anything that is lower than 99% or has less than 2 FLNC read support.

MAP AND COLLAPSE ISOFORMS

BENEFITS OF ISO-SEQ ANALYSIS APPLICATION

• High-quality transcripts
• Full-Length Non-concatemer reads
• Mapped & collapsed isoforms
• Removes artifacts
• Removes poly(A) tails

Iso-Seq Analysis Using pbBioConda

INSTRUCTIONS TUTORIAL

Follow the instructions tutorial for installing all the software needed.

• If you do not have an HPC server to install pbbioconda, you should have already:
  • Create an AWS account
  • Create an AWS Linux Instance to run Iso-Seq 3 Analysis Pipeline
  • Connect to your AWS Instance
• Upgrades and Install Software

DOWNLOAD THE DATA here (we will be using the data to do practice questions in the tutorial)

Example:

$ wget –nv https://downloads.pacbcloud.com/public/dataset/ISMB_workshop/isoseq3/alz.ccs.bam

SPECIFY ISO-SEQ PRIMERS

$ more primers.fasta

INPUT CCS BAM FILE

$ samtools view -h alz.ccs.bam

REFERENCE GENOME

$ grep '>’ hg38.fa # to list the headers per chromosome

SOFTWARE INSTALLATION CHECK

Access to your conda environment

$ source activate <name of your environment>

Check your installation

$ isoseq3 --version

$ lima --version

$ pbmm2 –-version

ISO-SEQ WORKFLOW

Click here if you want to look at the entire Iso Seq workflow on PacBio’s GitHub page

• Align to reference genome
• Remove redundancy

PRIMER REMOVAL & DEMULTIPLEXING

Command line:

lima --isoseq --dump-clips --peek-guess -j 24\

Input files:

• alz.ccs.bam #HiFi reads
• isoseq_primers.fasta #Iso-Seq primers

Output files:

• alz.demult.bam

Options:

• –isoseq #specialized isoseq option for lima
• –dump-clips # show the clipped primers
• –peek-guess # remove spurious false positive signal
• -j 24 # Number of threads to use

After completion, you will see the following files:

$ ls -ltrh

TRIMMING POLY(A) TAILS AND CONCATEMER REMOVAL

Command line:

isoseq3 refine --require-polya\

Input files:

• alz.demult.5p–3p.bam
• isoseq_primers.fasta

Output files:

• alz.flnc.bam

Options:

• –require-polya #if your transcripts have a polyA tail

After completion, you will see the following files:

$ ls -ltrh

ISOFORMS

Command line:

isoseq3 cluster alz.flnc.bam alz.polished.bam \
--verbose --use-qvs

Input files:

• alz.flnc.bam

Output files:

• alz.flnc.bam

Options:

• –verbose #if your transcripts have a polyA tail
• –use-qvs #Use CCS QVs, sets –poa-cov 100

After completion, you will see the following files:

$ ls -ltrh

Note: Because the ccs input is Polished, the isoseq3 cluster output is already polished!

MAP

Command line:

pbmm2 align hg38.fa alz.polished.hq.bam alz.aligned.bam
-j 24 --preset ISOSEQ –-sort --log-level INFO

Note: The pdf is wrong. It should be –sort, not -sort.

Input files:

• alz.polished.hq.bam
• hg38.fa

Output files:

• alz.aligned.bam

Options:

• -j 24 #Number of threads to use
• –preset ISOSEQ #select the alignment mode
• –sort #Generate sorted BAM file
• –log-level INFO #show progress

After completion, you will see the following files:

$ ls –ltrh

COLLAPSE

Command line:

isoseq3 collapse alz.aligned.bam alz.collapsed.gff

Input files:

• alz.aligned.bam

Output files:

• alz.collapsed.gff

After completion, you will see the following files:

$ ls –ltrh

PUBLICLY AVAILABLE ISO-SEQ DATA SETS here if you’re interested

ISO-SEQ ANALYSIS TERMINOLOGY

The pdf to this documentation can be found here

Iso-Seq pipeline with Bioconda and visualization using UCSC and IGV & Cupcake (hands on)

The Iso Seq Bioinformatics Tutorial we are using is here. See ‘1. isoseq3’ and answer the corresponding practice questions!

Note: We did not have time to do Cupcake during the workshop so try it out yourself and if you have questions feel free to reach out! See ‘2. Cupcake Fun’ and answer the corresponding practice questions!