MD Candidate @ USyd • Computational Biology & Software Engineering

Bridging clinical medicine, interpretable machine learning, and human-centric software.

I’m Jedrzej Konarkowski—a medical student at the University of Sydney with a background in computer science, biology, and translational bioinformatics research at Georgia Tech and Emory University School of Medicine. My work focuses on machine learning diagnostics in hematology, biobank-scale statistical genetics, and practical clinical software.

GitHubGoogle ScholarContact
Jedrzej Konarkowski
SOFTWARE & CLINICAL TOOLS

Flagship Systems & Applications

Deployed clinical pharmacology reference software and machine learning web applications for sickle cell diagnostics.

CLINICAL PHARMACOLOGY & PRODUCT2026

mypharm

Australian Clinical Drug Reference & Adherence Tracker

A comprehensive Australian clinical pharmacology reference and medication adherence platform designed to make drug monographs, PBS subsidy data, dosing regimens, and contraindication lookups fast and accessible for clinicians, students, and patients.

  • Tailored to Australian clinical practice with structured drug monographs, PBS subsidy status, Australian brand guides, and side-effect references.
  • Pairs rapid point-of-care medication library search with an interactive daily adherence and prescription scheduling tracker.
  • Live production web application engineered with modern TypeScript and a clean, accessible user interface.
Visit mypharm.info
TypeScriptReactClinical PharmacologyAustralian PBSDigital Health
mypharm.info
mypharm — Australian Clinical Drug Reference & Adherence Tracker interface
ACUTE HEMATOLOGY ML2025

SIGHT

Sickle Cell Identification from General Hematological Testing

A machine learning diagnostic screening tool designed for emergency medicine and acute triage. SIGHT predicts the likelihood of sickle cell disease from a single routine Complete Blood Count (CBC), helping clinicians flag undiagnosed or misclassified patients immediately without waiting for slow confirmatory electrophoresis.

  • Trained and validated across adult CBC cohorts combining NHANES and Grady Memorial Hospital clinical data.
  • Addresses acute care EHR discrepancies where confirmatory HPLC or electrophoresis is unavailable at triage.
  • Deployed as an interactive clinical Shiny web application alongside the ASH 2025 presentation.
Launch SIGHT WebappAbstract (Blood)
R / ShinyXGBoostClinical MLEmergency MedicineCBC Screening
GENOTYPE CLASSIFICATION2025

SHARP

Transfusion-Aware Hemoglobin Genotype Classification Pipeline

An interpretable, transfusion-aware machine learning framework that classifies specific sickle cell genotypes (HbSS, HbSC, HbSβ⁰, and HbSβ⁺) from routine hematological testing, trained against whole-genome sequencing (WGS) confirmed ground-truth cohorts.

  • Accounts for confounding recent blood transfusions (within 30 days) that frequently distort standard hemoglobin fraction interpretation.
  • Integrates SHAP-based feature attribution to provide transparent, clinician-auditable biomarker explanations.
  • Built on a retrospective cohort of 794 WGS-confirmed SCD patients and 298 HbAS controls.
Launch SHARP WebappAbstract (Blood)
R / ShinySHAPLightGBMWGS Ground-TruthHematology

Additional Lab & Engineering Projects

LEAP
Laboratory Information System & Clinical Data Pipeline

Modular full-stack data management system for wet-lab specimen tracking, hematology assay ingestion, and longitudinal clinical cohort harmonization.

PythonFlaskPostgreSQLBioinformatics
MouseHouse
Automated Vivarium Census & Computer Vision Tool

Edge-optimized object detection web application using YOLOv8 and OpenVINO to streamline non-invasive preclinical colony counting.

YOLOv8OpenVINOComputer VisionPython
LabGPT
Retrieval-Augmented Generation for Wet-Lab Protocols

Citation-grounded RAG assistant indexing institutional Standard Operating Procedures and experimental protocols for rapid benchside lookup.

PythonRAGVector DBLLMs
DEFORM
Sickle Cell Pathophysiology Strategy Simulation

Pixel-art tower defense game built in Godot modeling microvascular vaso-occlusion, hypoxia gradients, and hematological therapeutics to educate about Sickle Cell Disease pathophysiology and clinical work.

GodotGDScriptMedical EducationGame Dev
RESEARCH & SCHOLARSHIP

Publications & Conference Abstracts

Statistical genetics in the UK Biobank and translational machine learning & hemorheology research in sickle cell disease.

Google Scholar Profile
Peer-Reviewed Journal Publications
BMC Medical GenomicsBMC Med Genomics 18, 163 (2025)

Assessment of genetic and metabolite associations of branched chain amino acids with metabolic disease in the UK Biobank using Mendelian randomization.

Konarkowski, J., Astore, C. & Gibson, G. (2025). BMC Med Genomics 18, 163 (2025).

Integrative Phenome-Wide Association Study (PheWAS) and Mendelian randomization investigation across the UK Biobank evaluating plasma leucine, isoleucine, and valine against polygenic risk to untangle directional causality across hypertension, type 2 diabetes, obesity, and chronic kidney disease.

First AuthorUK BiobankMendelian RandomizationPheWASMetabolomics
Publication (DOI)Google Scholar
Conference Presentations & Abstracts
Blood (ASH 2025)Blood, 146, 1177 (2025)

SIGHT: Sickle cell identification from general hematological testing, a machine learning approach for rapid screening in emergency medicine.

Konarkowski, J., Wade, J., & Sheehan, V. (2025). Blood, 146, 1177 (2025).

First-author machine learning screening framework trained on routine Complete Blood Count (CBC) parameters to rapidly identify sickle cell disease in acute and emergency care settings without waiting for confirmatory electrophoresis.

First AuthorSIGHTEmergency MedicineClinical ML
Launch SIGHT WebappAbstract (Blood)
Blood (ASH 2025)Blood, 146, 4743 (2025)

High-accuracy, transfusion-aware machine learning pipeline for hemoglobin genotype classification with SHAP-based interpretability.

Wade, J., Konarkowski, J., Bhat, V., Agbakwuru, U., & Sheehan, V. (2025). Blood, 146, 4743 (2025).

Transfusion-aware classification pipeline (SHARP) trained on whole-genome sequencing confirmed cohorts to differentiate hemoglobinopathy genotypes (HbSS, HbSC, HbSβ⁰, HbSβ⁺) from routine hematological testing with SHAP feature attribution.

Co-AuthorSHARPSHAP InterpretabilityHematology
Launch SHARP WebappAbstract (Blood)
Blood (ASH 2025)Blood, 146, 7855 (2025)

Higher baseline whole blood viscosity is associated with chronic SCD complications.

Patwardhan, A., Patel, A., Konarkowski, J., Wade, J., & Sheehan, V. (2025). Blood, 146, 7855 (2025).

Clinical rheology investigation demonstrating how elevated baseline whole blood viscosity stratifies longitudinal risk for chronic organ complications in sickle cell disease.

Co-AuthorHemorheologyBlood ViscositySCD
Abstract (Blood)
Blood (ASH 2025)Blood, 146, 1152 (2025)

Red cell rheology predicts acute chest syndrome and vaso-occlusive events in sickle cell disease.

Patel, A., Agbakwuru, U., Patwardhan, A., Wade, J., Konarkowski, J., Rivera, M., Yoo, J., & Sheehan, V. (2025). Blood, 146, 1152 (2025).

Translational modeling of functional erythrocyte rheology biomarkers to predict vaso-occlusive crises and acute chest syndrome trajectories.

Co-AuthorRed Cell RheologyAcute Chest SyndromeVaso-Occlusion
Abstract (Blood)
Awards & Honours
  • 2025 Robert A. Pierotti Memorial Scholarship
    Georgia Institute of Technology, March 2025
  • 1st Place, Annual Conference of Psychological Research
    Georgia Institute of Technology, April 2024