Tortoise: de novo variant calling workflow, CPU version

Developed and maintained by Jeffrey Ng

Washington University in St. Louis Medical School

Tychele N. Turner, Ph.D., Lab

This pipeline takes in aligned, short-read Illumina data, in the form of .cram or .bam files, for a family trio and reports de novo variants in a .vcf file. This is a modified version pipeline run from Ng et al. 2022, making use of the open-source, CPU-only versions of the programs outlined in the paper. This pipeline uses DeepVariant (Poplin et al. 2018) and GATK Haplotypecaller to call variants (Poplin et al. 2017), GLnexus (Lin et al. 2018) for joint-genotyping, and lastly, a custom workflow to call de novo variants.

How to Run

Input

Two main inputs:

1. .bam or .cram files for the trio(s)
2. A comma-delimited text file, with one trio per line, with sample IDs formatted in the following way: Father,Mother,Child

You may run WGS, WES, and PacBio long-read sequencing data.

You may download the full 30x WGS NA12878 trio .cram files from the 1000 Genomes Project to test the workflow, links are found below:

wget -q ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR323/ERR3239334/NA12878.final.cram
wget -q ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR323/ERR3239334/NA12878.final.cram.crai
wget -q ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR398/ERR3989341/NA12891.final.cram
wget -q ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR398/ERR3989341/NA12891.final.cram.crai
wget -q ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR398/ERR3989342/NA12892.final.cram
wget -q ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR398/ERR3989342/NA12892.final.cram.crai

Alternatively, small example .bam files can be found in the test_code folder for faster testing purposes. These .bam files were created from the 30x .cram files above, isolating a small region on chromosome 5.

To download the GRCh38 reference, please use the following:

wget -q http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/GRCh38_reference_genome/GRCh38_full_analysis_set_plus_decoy_hla.fa
wget -q http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/GRCh38_reference_genome/GRCh38_full_analysis_set_plus_decoy_hla.fa.fai
wget -q http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/GRCh38_reference_genome/GRCh38_full_analysis_set_plus_decoy_hla.dict
 

If you would like to download the RepeatMasker files, please use the following links:

Centromeres

wget -q https://de.cyverse.org/dl/d/B42A0F3D-C402-4D5F-BBD5-F0E61BE2F4AC/hg38_centromeres_09252018.bed.gz
wget -q https://de.cyverse.org/dl/d/37B13DB5-0478-4C4B-B18D-33AFB742E782/hg38_centromeres_09252018.bed.gz.tbi

LCR plus 5 bp buffer

wget -q https://de.cyverse.org/dl/d/870755FF-CD04-4010-A1EC-658D7E1151EF/LCR-hs38-5bp-buffer.bed.gz
wget -q https://de.cyverse.org/dl/d/01D038EA-51CC-4750-9814-0BB3784E808E/LCR-hs38-5bp-buffer.bed.gz.tbi

Recent repeats plus 5 bp buffer

wget -q https://de.cyverse.org/dl/d/185DA9BC-E13D-429B-94EA-632BDAB4F8ED/recent_repeat_b38-5bp-buffer.bed.gz
wget -q https://de.cyverse.org/dl/d/4A6AF6EF-D3F0-4339-9B8E-3E9E83638F00/recent_repeat_b38-5bp-buffer.bed.gz.tbi

CpG Locations

wget -q https://de.cyverse.org/dl/d/786D1640-3A26-4A1C-B96F-425065FBC6B7/CpG_sites_sorted_b38.bed.gz
wget -q https://de.cyverse.org/dl/d/713F020E-246B-4C47-BBC3-D4BB86BFB6E9/CpG_sites_sorted_b38.bed.gz.tbi

Running

This pipeline can be run using the Docker image found here: tnturnerlab/tortoise:v1.2 We also provide the Dockerfile if you would like to make modifications.

Snakemake

Setting up the config.json file

Before running, please make any necessary changes to these options below in the config.json.

• regions: "/region" If you don't have the RepeatMasker files, please make this entry blank
• gq_value: 20 gq value filter
• depth_value: 10 Depth value filter
• suffix: "_supersmall.bam" Suffix of your data files. Assumes input files are <sample_name><suffix>
• family_file: "/dnv_wf_cpu/<your_family_file>"
• chrom_length: Optional chromosome length file, use if you are not using human reference GRCh38. Can leave it blank if using GRCh38. Please make this a two-column, tab-delimited file, with the first chromosome and the second column the length of the chromosome
• "glnexus_dv_model": Optional model file, if you are running Deepvariant on WES data, please add "DeepVariantWES". Otherwise, leave this blank
• "interval_file": If you are using an interval, please provide the path here. If not, please leave this blank. ,
• "dv_model": Please add the DeepVariant model here, defaults to WGS. Change this to WES or PACBIO if you are running those types of data

Running on a LSF server

For ease of use, we recommend having as much of the data in the same head directory as possible.

First, set up the LSF_DOCKER_VOLUMES:

export LSF_DOCKER_VOLUMES="/path/to/crams/:/data_dir /path/to/reference:/reference /path/to/this/git/repo/:/dnv_wf_cpu/ /path/to/RepeatMasker/files:/region"

Then run this command:

bsub -q general -oo %J.main.log -R 'span[hosts=1] rusage[mem=5GB]' -a 'docker(tnturnerlab/tortoise:v1.2)' /opt/conda/envs/snake/bin/snakemake -s /dnv_wf_cpu/tortoise_1.2.smk -j 6 --cluster-config cluster_config.json --cluster "bsub -q general -R 'span[hosts=1] rusage[mem={cluster.mem}]' -n {cluster.n} -M {cluster.mem} -a 'docker(tnturnerlab/tortoise:v1.2)' -M {cluster.mem} -oo %J.log.txt" -k --rerun-incomplete -w 120 

Running locally

To run the small test samples provided, please ensure you have at least 16GB of RAM. To run a full-sized trio, please ensure your machine has at least 64GB of RAM. The pipeline will use as many CPUs as possible.

For ease of use, we recommend having as much of the data in the same head directory as possible.

To run this locally, please run this command:

docker run -v "/path/to/crams/:/data_dir" -v "/path/to/hare/code:/dnv_wf_cpu" -v "/path/to/data:/data"  tnturnerlab/tortoise:v1.2 /opt/conda/envs/snake/bin/snakemake -s /dnv_wf_cpu/tortoise_1.2.smk -j 6 --cores -k --rerun-incomplete -w 120 

Cromwell workflow

We also provide this workflow in a .wdl format. Unlike the Snakemake, you will be able to run Parabricks directly from this workflow, instead of separately. You can also run this workflow in the cloud. To run this, you'll need to download the Cromwell .jar found here. This wdl was specifically tested on cromwell-83.

The basic config file looks like this:

{
  "jumping_tortoise.glnexus_cpu": "Int (optional, default = 32)",
  "jumping_tortoise.glnexus_ram_hc": "Int (optional, default = 250)",
  "jumping_tortoise.maxPreemptAttempts": "Int (optional, default = 3)",
  "jumping_tortoise.extra_mem_hc": "Int (optional, default = 65)",
  "jumping_tortoise.test_intersect": "File", #pathway to the test_intersect.py file
  "jumping_tortoise.chrom_length": "File? (optional)",  #Optional chromosome length file if you are not using Human build GRCh38
  "jumping_tortoise.combinedAndFilter.extramem_GLDV": "Int? (optional)",
  "jumping_tortoise.pathToReference": "File",  #pathway to tarball of reference information
  "jumping_tortoise.glnexus_ram_dv": "Int (optional, default = 100)",
  "jumping_tortoise.wes": "Boolean (optional, default = false)",   #Please set this to true if you are analyzing WES data
  "jumping_tortoise.glnexus_DV.extramem_GLDV": "Int? (optional)",
  "jumping_tortoise.hare_docker": "String (optional, default = \"tnturnerlab/hare:v1.1\")",
  "jumping_tortoise.cram_files": "Array[Array[WomCompositeType {\n cram -> File\ncrai -> File \n}]]",  #cram/bam file input, please see example for formatting
  "jumping_tortoise.regions": "File? (optional)",  #This is the tarball for your RepeatMaster files
  "jumping_tortoise.cpu_dv": "Int (optional, default = 32)",
  "jumping_tortoise.naive_inheritance_trio_py2": "File",
  "jumping_tortoise.sample_suffix": "String",  #suffix of the input cram file.  If your sample was NA12878.final.cram, you would put ".final.cram" here
  "jumping_tortoise.num_ram_dv": "Int (optional, default = 120)",
  "jumping_tortoise.glnexus_HC.extramem_GLDV": "Int? (optional)",
  "jumping_tortoise.deep_docker": "String (optional, default = \"tnturnerlab/tortoise:v1.1\")",
  "jumping_tortoise.filter_glnexuscombined_updated": "File",
  "jumping_tortoise.gq": "Int (optional, default = 20)",
  "jumping_tortoise.extra_mem_dv": "Int (optional, default = 65)",
  "jumping_tortoise.trios": "Array[WomCompositeType {\n father -> String\nmother -> String\nchild -> String \n}]",   #trios MUST be in the same order as trios in cram_file
  "jumping_tortoise.depth": "Int (optional, default = 10)",
  "jumping_tortoise.deep_model": "String (optional, default = \"WGS\")",  #Please change this to WES or PACBIO depending if you are running those types of data
  "jumping_tortoise.num_ram_hc": "Int (optional, default = 64)",
  "jumping_tortoise.interval_file": "String (optional, default = \"None\")",
  "jumping_tortoise.glnexus_deep_model": "String (optional, default = \"DeepVariant\")",
  "jumping_tortoise.reference": "String",
  "jumping_tortoise.cpu_hc": "Int (optional, default = 4)"
} 

Required arguments are highlighted in the comments above. We have provided an example config to help with formatting. Please modify the computational requirements to fit your HPC. If you are going to use this wdl, please tarball your reference files. If you are running WES data, please include your capture region in this tarball.

tar -jcf reference.tar.bz2 reference.fa reference.fa.fai reference.dict

You will also need to put the RepeatMaster files into a separate tarball.

Please find the Cromwell documentation for a submission command that fits your specific HPC, but generally it would be run like this:

java -jar cromwell-83.jar run tortoise_1.2.wdl --inputs test_wdl_config.json

Output

With the regions option

After the run, you'll find two main output files found in the folder called out:

• <child_name>.glnexus.family.combined_intersection_filtered_gq_<gq_value>_depth_<depth_value>_position.vcf
  • This file holds the de novo variants
• <child_name>.glnexus.family.combined_intersection_filtered_gq_<gq_value>_depth_<depth_value>_position_all.vcf
  • This file holds the de novo variants specifically within CpG regions.

Without the region option

Your main output file is:

• <child_name>.glnexus.family.combined_intersection_filtered_gq_<gq_value>_depth_<depth_value>.vcf
  • This file holds de novo variants that are not filtered by regions mentioned above. The other two files will just contain the header but otherwise will be empty.

When running the small test bams, you will find these two DNVs in both NA12878.glnexus.family.combined_intersection_filtered_gq_20_depth_10.vcf or NA12878.glnexus.family.combined_intersection_filtered_gq_20_depth_10_position.vcf:

chr5 	51158671    	chr5_51158671_A_G A      	G     	43    	.        	AC=1;AF=0.167;AN=6;INH=denovo_pro;TRANSMITTED=no;set=Intersection  	GT:AD:DP:GQ:PL:RNC     	0/1:20,10:30:44:43,0,58        	0/0:26,0:26:50:0,123,1229 	0/0:19,0:19:50:0,81,809
chr5 	52352927    	chr5_52352927_T_C  T      	C     	56    	.        	AC=1;AF=0.167;AN=6;INH=denovo_pro;TRANSMITTED=no;set=Intersection  	GT:AD:DP:GQ:PL:RNC     	0/1:17,15:32:55:56,0,61        	0/0:28,0:28:50:0,102,1019 	0/0:26,0:26:50:0,114,1139

If you want to use the WES filter, please modify the config.json file in the wes_filtering folder. You can run the Snakemake with a docker command like this:

docker run -v "/path/to/script:/wes_filter" -v "/path/to/data:/data"  tnturnerlab/tortoise:v1.2 /opt/conda/envs/snake/bin/snakemake -s wes_filter.smk --cores -s /wes_filter/wes_filter.smk --cores 

This script will separate your WES DNVs into high and low confidence files. One as a bed file and another as a tab-delimited .txt file.

Software Requirements

• All software requirements are built into the docker image. The software packages built are as follows:
  • DeepVariant v1.4
  • GATK v4.2.0.0
  • GLnexus v1.4.1
  • Snakemake v7.15.2-0
  • python 3.9.7
  • python 2.7
  • BCFtools v1.11
  • SAMtools v1.11
  • bedtools v2.29.2
  • tabix v1.11
  • vcflib v1.0.0-rc0
  • pybedtools v0.9.0