LOng-read Multi-omics PipelinE

A pipeline for aligning long read pacbio or nanopore fastq files, calling SNVs, SVs and methylation, with phasing and filtering.

QC:

• Picard
• fastQC

Alignment:

• minimap2

SNV calling:

• deepvariant
• bcftools (for lower quality data)

SV calling:

• Sniffles v1

CNV calling:

• CNV pythor

Expansion repeats:

• TRGT (https://github.com/PacificBiosciences/trgt)

Phasing:

• WhatsHap

Methylation:

• PB: pb-cpg-tools
• ONT: specify pore; for < R10 nanopolish, R10

Workflow: #image / clean fastq -> bam (minimap2) snv calling and annotation (bcftools) phasing and haplotagging of bam file (whatshap) methylation calling (ont) (nanopolish) filtering SVs | extraction of methylation info (SVDB | nanopolsih/cpgtools)

Install

Dependencies:

Nextflow version 22.10.2

python3

Docker or singularity

samtools

bcftools

FastQ

For ONT methylation calling: hdf5

git clone < repo >

edit config to your needs

RUN

nextflow run kristinebilgrav/lompe 

REQUIRED:
-c < config > 
--input < ONT: folder containing 'fastq_pass' with fastq files/ samplesheet with folder paths PB: samplesheet with paths to bam or fastq file  > 
--output < pathto/output/folder> will generate pathto/output/folder/SampleID_out
--style < ont OR pb>   
--file < 'fastq' OR 'bam' OR 'ubam'>

OPTIONAL:
-with-trace < creates log file >
--bcftools < if for some reason want to use bcftools instead of deepvariant >

the samplesheet.csv need to contain a header: SampleID,SamplePath where rows follow with sampleid,folder/orbam/path

for PB fastq files a path to the fastq file (gz) is expected, OR a bam file with methylation tags incorported

for ONT samples, a folder called 'fast5_pass' is expected in the same directory as the 'fastq_pass' folder OR path to folder with ubams to remap to one

Input: nanopore: folder containing fastq_pass and fast5_pass folders with fastq.gz / fast5 OR path to folder containing ubams pacbio: folder containing fastq_pass with fastq.gz OR bam file

Methylation annotation of bam: Only performed on ONT samples.

For SVDB database annotation and filtering: Local SVDB database with variants to use for filtering

The config file needs to be edited with path to reference genome, containers / paths to program, clusteroptions, annotation databases. example config file can be found as LOMPE.config

Output

ONT: 
sample_id.phased.haplotagged.bam(.bai) HP and Mm tags 
sample_id.snv.filtered.vcf
sample_id.output.VEP.query.filtered.vcf
sample_id.methylsites.tsv with ref positions, read name and log_likihood of methylation (nanopolish)
sample_id.methylfrequency.tsv with log_lik_ratio summarized for each position (nanopolish)
sample_id.wgsmetrics.txt
sample_id.fastQC.out/

PB: 
sample_id.phased.bam(.bai)
sample_id.snv.filtered.vcf
sample_id.output.VEP.query.filtered.vcf
sample_id.wgsmetrics.txt
sample_id.fastQC.out/

if bam with methyl sites:
sample_id.combined/hap1/hap2.denovo.bed pb-cpg-tools for more info