allspise

Manuscript is in preparation

Introduction

allspise is a pipeline for identifying allele-specific isoform expression using long read sequencing.

This pipeline was developed to investigate allelic expressions on the isoform level. Allspise was built using Nextflow DSL2 that can run tasks across multiple compute infrastructures in a very portable manner. It uses Docker/Singularity containers, making installation trivial and results highly reproducible.

Documentation

For details of usage and output interpretation see docs.

Quick Start

1. Install Nextflow (>=21.04.0)

2. Install Docker. If you are not able to use containers, see the local installation instruction.

3. Download the pipeline and test it on a minimal test dataset with a single command:

   nextflow main.nf -profile test,docker
   
   # without docker 
   # please install local dependencies following the instruction
   nextflow main.nf -profile test

   If it runs successfully, you'll get:

   [e9/b78794] process > ALLSPISE:INPUT_CHECK:SAMPLESHEET_CHECK (samplesheet.csv)     [100%] 1 of 1 ✔
   [f8/f54328] process > ALLSPISE:SCISORSEQR_PROCESSES:GRAB_STUFF (test)              [100%] 1 of 1 ✔
   [8f/8303b2] process > ALLSPISE:SCISORSEQR_PROCESSES:MAP_AND_FILTER_MINIMAP2 (test) [100%] 1 of 1 ✔
   [3c/8a846b] process > ALLSPISE:SCISORSEQR_PROCESSES:LONG_READ_INFO (test)          [100%] 1 of 1 ✔
   [e6/9f7b9a] process > ALLSPISE:VARIANT_CALLING:RUN_LONGSHOT (test)                 [100%] 1 of 1 ✔
   [3b/5eb082] process > ALLSPISE:VARIANT_CALLING:READ_TO_SNP (test)                  [100%] 1 of 1 ✔
   [41/d44ae0] process > ALLSPISE:DIFFERENTIAL_ANALYSIS:DIFFERENTIAL_TEST (test)      [100%] 1 of 1 ✔
   

   The results obtained from the mini test dataset can be viewed in the test_results/, soon after you execute the above command ;).

4. Start running your own analysis! E.g.:

   nextflow main.nf -profile docker \
       --samplesheet samplesheet.csv \
       --ref_genome /path/to/ref.fa.gz \
       --ref_chroms /path/to/ref_chroms/ \
       --gtf /path/to/gencode.gtf.gz \
       --cage_bed /path/to/cage_peaks.bed.gz \
       --polyA_bed /path/to/polya.bed.gz \
       --aligner STARlong \
       --star_index /path/to/star_index \
       --var_caller IsoPhase

Pipeline Summary

The allspise workflow encompasses 4 sub-workflows:

1. INPUT_CHECK: check the format of samplesheet and whether input files exist, etc.
2. SCISORSEQR_PROCESSES: align long reads and extract full-length transcripts using R package scisorseqr
3. VARIANT_CALLING: call SNPs and conduct relevant analysis. For now we only include two callers: IsoPhase and Longshot; for method details of SNP analysis see this paper
4. DIFFERENTIAL_TEST: perform differential tests to identify allele-specific isoforms/TSS/polyA-sites; for method details see this paper

Flowchart: