TOMM40_WGS: Genotyping TOMM40'523 Poly-T Polymorphisms from Whole-Genome Sequencing

TOMM40_WGS is a Snakemake pipeline that enables automated genotyping of the TOMM40'523 poly-T repeat from WGS data. It integrates STR genotyping tools, k-mer analysis, and machine learning to provide accurate repeat length and genotype predictions.

---

Background

The TOMM40'523 poly-T repeat polymorphism (rs10524523) is a variable-length poly-T repeat located in intron 6 of the TOMM40 gene on chromosome 19. This polymorphism has been associated with age of onset of Alzheimer's disease (AD) and cognitive decline.

The repeat length is typically categorized into three classes:

• Short (S): ≤19 T residues
• Long (L): 20-29 T residues
• Very Long (VL): ≥30 T residues

---

🚀 Quick Start

Software Dependencies

• Conda: install via Miniforge (recommended)

Install with bioconda

conda install -c conda-forge -c bioconda tomm40_wgs

Use the TOMM40_WGS launcher script to run the pipeline with minimal setup.

Alternatively: Create a conda environment

conda config --set channel_priority strict
conda create -y -n TOMM40_WGS_env -c conda-forge -c bioconda \
snakemake==9.3.0 samtools==1.21 pandas==2.2.3
conda activate TOMM40_WGS_env

Clone the repository and install dependencies

git clone https://github.com/RushAlz/TOMM40_WGS.git
cd TOMM40_WGS
chmod u+x TOMM40_WGS

# Get help
./TOMM40_WGS --help

---

Basic usage (recommended)

# Basic usage with a single BAM file (aligned to GRCh38 by default)
TOMM40_WGS --input_wgs sample.bam --ref_fasta GRCh38.fa --cores 2

# Basic usage with multiple BAM files using glob pattern (quotes are required)
TOMM40_WGS --input_wgs "*.bam" --ref_fasta GRCh38.fa --cores 2

# Basic usage with CRAM files 
TOMM40_WGS --input_wgs "*.cram" --ref_fasta GRCh38.fa --genome_build GRCh38 --cores 2

# Using a sample table (TSV file with sample_id and path columns)
TOMM40_WGS --input_table samples.tsv --ref_fasta GRCh38.fa --cores 2

# With a custom configuration file
TOMM40_WGS --configfile custom_config.yaml --cores 2

This script wraps Snakemake, applies default values when needed, and accepts any extra Snakemake flags (e.g., --dryrun).

---

End-to-end testing with real data (data from 1000 Genomes Project)

Reference for the data

Install Miniforge3

# Skip if conda is already installed
wget -O Miniforge3.sh "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
bash Miniforge3.sh -b -p "${HOME}/conda"
source "${HOME}/conda/etc/profile.d/conda.sh"

Create TOMM40_WGS_env

# Skip if already created
conda config --set channel_priority strict
conda create -y -n TOMM40_WGS_env -c conda-forge -c bioconda \
snakemake==9.3.0 samtools==1.21 pandas==2.2.3
conda activate TOMM40_WGS_env

Clone the repository and create a testing directory

git clone https://github.com/RushAlz/TOMM40_WGS.git
chmod u+x TOMM40_WGS/TOMM40_WGS

# Make a temporary working directory
mkdir -p tomm40_wgs_test
cd tomm40_wgs_test

Test TOMM40_WGS with GRCh38 CRAMs

# Download the reference genome
REF_FASTA="https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/GRCh38_reference_genome/GRCh38_full_analysis_set_plus_decoy_hla.fa"
REF_FASTA_LOCAL=$PWD/"GRCh38_full_analysis_set_plus_decoy_hla.fa"
wget -O ${REF_FASTA_LOCAL} ${REF_FASTA}
wget -O ${REF_FASTA_LOCAL}.fai ${REF_FASTA}.fai

# Subset the CRAM file to the TOMM40 region
CRAMFILE_1="https://ftp.sra.ebi.ac.uk/vol1/run/ERR324/ERR3240114/HG00096.final.cram"
CRAM_1_LOCAL=$PWD/"HG00096.GRCh38.cram"
samtools view -h -T GRCh38_full_analysis_set_plus_decoy_hla.fa -C ${CRAMFILE_1} "chr19:44399792-45399826" > ${CRAM_1_LOCAL} 
samtools index ${CRAM_1_LOCAL}

# Download a second CRAM file for testing
CRAMFILE_2="https://ftp.sra.ebi.ac.uk/vol1/run/ERR323/ERR3239334/NA12878.final.cram"
CRAM_2_LOCAL=$PWD/"NA12878.GRCh38.cram"
samtools view -h -T GRCh38_full_analysis_set_plus_decoy_hla.fa -C ${CRAMFILE_2} "chr19:44399792-45399826" > ${CRAM_2_LOCAL}
samtools index ${CRAM_2_LOCAL}

# Run the pipeline specifying the input files as a glob (quotes are required) 
../TOMM40_WGS/TOMM40_WGS --input_wgs "*.cram" \
--ref_fasta GRCh38_full_analysis_set_plus_decoy_hla.fa \
--output_dir results_GRCh38 \
--cores 2

# Alternatively, run the pipeline specifying the input files as a TSV file
echo -e "sample_id\tpath\nHG00096\t${CRAM_1_LOCAL}\nNA12878\t${CRAM_2_LOCAL}" > samples.tsv
../TOMM40_WGS/TOMM40_WGS --input_table samples.tsv \
--ref_fasta GRCh38_full_analysis_set_plus_decoy_hla.fa \
--output_dir results_GRCh38 \
--cores 2

# Check results
cat results_GRCh38/predictions/tomm40_predictions_summary.tsv | column -t

Test TOMM40_WGS with b37 BAMs

# Download the reference genome
REF_FASTA="https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/phase2_reference_assembly_sequence/hs37d5.fa.gz"
REF_FAI="https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/phase2_reference_assembly_sequence/hs37d5.fa.gz.fai"
wget ${REF_FASTA}
wget ${REF_FAI}
REF_FASTA_LOCAL=$PWD/"hs37d5.fa.gz"

# Subset the BAM file to the TOMM40 region
BAMFILE_1="https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/phase3/data/HG00096/high_coverage_alignment/HG00096.wgs.ILLUMINA.bwa.GBR.high_cov_pcr_free.20140203.bam"
BAM_1_LOCAL=$PWD/"HG00096.hg37.bam"
samtools view -h -b ${BAMFILE_1} "19:44903049-45903083" > ${BAM_1_LOCAL}.tmp
samtools addreplacerg -r '@RG\tID:HG00096\tSM:HG00096\tLB:lib1' -o ${BAM_1_LOCAL} ${BAM_1_LOCAL}.tmp
samtools index ${BAM_1_LOCAL}

# Download a second CRAM file for testing
BAMFILE_2="https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/phase3/data/NA12878/high_coverage_alignment/NA12878.mapped.ILLUMINA.bwa.CEU.high_coverage_pcr_free.20130906.bam"
BAM_2_LOCAL=$PWD/"NA12878.hg37.bam"
samtools view -h -b ${BAMFILE_2} "19:44903049-45903083" > ${BAM_2_LOCAL}.tmp
samtools addreplacerg -r '@RG\tID:NA12878\tSM:NA12878\tLB:lib1' -o ${BAM_2_LOCAL} ${BAM_2_LOCAL}.tmp

samtools index ${BAM_2_LOCAL}

# Run the pipeline specifying the input files as a glob (quotes are required) 
../TOMM40_WGS/TOMM40_WGS --input_wgs "*.bam" \
--ref_fasta hs37d5.fa.gz \
--genome_build b37 \
--output_dir results_b37 \
--cores 2 


# Check results
cat results_b37/predictions/tomm40_predictions_summary.tsv | column -t

---

Pipeline Overview

The pipeline performs the following steps for each input sample:

1. Subset CRAM/BAM to the TOMM40 region
2. STR genotyping using:
   • HipSTR
   • GangSTR
   • ExpansionHunter
3. K-mer extraction from poly-T region using Jellyfish
4. Feature processing using R scripts
5. ML-based prediction of repeat lengths and TOMM40 genotype class
6. Aggregate predictions into a summary table

---

Input Configuration

Input data can be specified in two ways:

1. Using command line parameters:

   • --input_wgs: A single file, a glob pattern (must be quoted, e.g., "*.cram")
   • --input_table: Path to a TSV file with sample_id and path columns

2. Using configuration file:

   • input_glob: A glob pattern to CRAM/BAM files
   • input_table: Path to a TSV file with sample_id and path columns

When both --input_wgs and --input_table are specified, --input_table takes precedence.

input_table format:

sample_id    path
sample1      /path/to/sample1.bam
sample2      /path/to/sample2.bam

• genome_build: Specify the genome build according to the reference genome used for alignment. The default is GRCh38.

Supported builds include:

• GRCh38
• GRCh37
• hg19
• b37

# TOMM40`523 (rs10524523) coordinates mapping to genome builds
"GRCh38": "chr19:44899792-44899826"
"GRCh37": "chr19:45403049-45403083"
"hg19": "chr19:45403049-45403083"
"b37": "19:45403049-45403083"

Important currently only GRCh38 has been extensively tested.

---

Outputs

Each sample generates:

• results/GangSTR/, HipSTR/, ExpansionHunter/: Genotyping VCFs for each STR tool
• results/jellyfish/{sample}.polyT_kmer.txt: k-mer counts
• results/features/{sample}_features.txt: Combined features for ML
• results/predictions/{sample}.tomm40_prediction.txt: Length/genotype predictions for {sample}
• results/predictions/tomm40_predictions_summary.tsv: Combined summary

---

License

TOMM40_WGS is licensed under the GPL-2.0 License - see the LICENSE file for details. Important licensing information is available here.

---

Citation

If you use TOMM40_WGS, please cite:

Vialle RA et al. (2025). Genotyping TOMM40'523 poly-T polymorphisms using whole-genome sequencing. Human Genetics and Genomics Advances. https://doi.org/10.1016/j.xhgg.2025.100488