Dual-caller strategy
GATK and DeepVariant calls are merged with Spark/Glow, reaching 0.99 F1/recall on GIAB benchmarks.
Bioinformatician · Medical Geneticist
I build production-grade genomics pipelines that turn raw sequencing reads into ▍
I started out in a hematology lab, running blood tests for patients with leukemia. Today I write the software that reads their genomes.
I trained in Medical Laboratory Sciences and Human Genetics, and my work now sits where clinical genetics meets software engineering. I design scalable Nextflow pipelines, benchmark variant callers against GIAB truth sets, and write clinical reports that help families get a diagnosis.
My research covers hereditary spastic paraplegia, rare neurological disorders, cancer transcriptomics, and endogenous retroviruses in Parkinson’s disease at single-cell resolution.
##fileformat=VCFv4.3
##source=fardad.tech
##INFO=<ID=ROLE,Description="Current calling">
#CHROM POS ID REF ALT
chrIR 2015 fardad MedLab Bioinformatics
QUAL FILTER
99 PASS
INFO
ROLE=Bioinformatician,Medical_Geneticist;
FOCUS=Rare_Disease,Cancer_Genomics,scRNA;
STACK=Nextflow,GATK,DeepVariant,Spark;
LANG=Python,Rust,R,Bash,Groovy;
ACMG=Pathogenic # for bad pipelines
The Nextflow (DSL2) suite I designed and built at Genomist Ltd. It is containerized, versioned and benchmarked, and it runs in clinical production.
GATK and DeepVariant calls are merged with Spark/Glow, reaching 0.99 F1/recall on GIAB benchmarks.
Parallel processing brings each exome under one hour, fast enough for high-throughput cohorts.
HPO-integrated CNV calling plus cancer biomarker modules for TMB, MSI, ROH and LOH.
Every module is pinned with Docker and DSL2 versioning. DRAGMAP replaces BWA to resolve complex regions better.
Genomist Ltd. · Tehran
BOKU University, Vienna · Remote
Shahid Sadoughi University of Medical Sciences · Remote
MOM Fertility Clinic · Remote
Salvia Medical Sciences · CytoCity · Royan Institute
Zagross Medical Laboratory · Kermanshah
University of Social Welfare and Rehabilitation Sciences, Tehran
Thesis: whole-exome sequencing in five Iranian families with hereditary spastic paraplegia to find disease-causing variants. Supervised by Dr. Afagh Alavi and Prof. Hossein Najmabadi.
Kermanshah University of Medical Sciences
Full scholarship recipient.
Rare neurological disease genetics, cancer genomics and single-cell sequencing.
Current Journal of Neurology
↗ Aug 2021Canadian Journal of Neurological Sciences
↗ Jun 2021Canadian Journal of Neurological Sciences
↗ Feb 2021American Journal of Medical Genetics Part A
↗I built a PacBio HiFi pipeline for small and structural variants, tandem repeat expansions, long-range phasing and native DNA methylation analysis.
Explore ClinLR documentation ↗I built a custom composite reference of the human genome plus known HERVs to detect endogenous retroviruses in single-cell data from several Parkinson’s cohorts. This is a collaboration with Dr. Michael Lappe at BRIC, University of Copenhagen.
Seurat-based scRNA-seq analysis of GSE103224 showing adipocyte accumulation and weakened immune responses in IDH1 R132H tumors.
Presented at the 4th International and 16th Iranian Genetics Congress.
European Human Genetics Conference 2022.
A systematic review using network analysis of clinical findings. Presented with a companion talk on HSP heterogeneity.
Shown as an expression heatmap. Brighter cells mean deeper experience.
Germline & somatic variant calling · CNV · STR · PacBio long-read · WES · scRNA-seq · bulk RNA-seq · ChIP-seq · GIAB validation
Nextflow DSL2 · Docker / Singularity · scatter/gather · Spark / Glow · SLURM & HPC
ACMG classification · HPO integration · TMB / MSI / ROH / LOH · clinical reporting
Model development & deployment · RAG · LLM fine-tuning · agentic AI · MLOps
DISCOVER PhD Scholarship, H2020 Marie Skłodowska-Curie (2022)
Best Poster Award, Iranian Genetics Congress (2020)
Out of 65,333 applicants, M.Sc. entrance exam (2019)
Fundamental Skills in Bioinformatics, plus courses from Google, Imperial, Fred Hutch and IBM