Greg Starek Email & Phone Number

AI and statistical modeling to improve health outcomes

San Francisco, California, United States

• + Built predictive and descriptive models to aid in player evaluation at scale + Led development of new biomechanics software and data pipeline, transforming raw data into actionable insights+ Built API to enable easy.
• Tensorflow
• scikit-learn
• xgboost
• openCV
• flask

Carlsbad, CA

• Tools:
• Tensorflow
• scikit-learn
• PyMC3
• Docker
• Postgres, MySQL, PySpark

Madison, WI

• Analyzed biophysical effects of genetic mutation linked to sudden cardiac arrest in eight year old
• Developed novel cardiac safety assay to assess risk of drug induced arrhythmias, allowing more drugs to safely reach the marketplace
• Identified antibodies capable of modulating hERG action potential current, revealing new sites for potential cardiac drug binding

Remote Session

• Developed web app to identify Instagram users suspected of purchasing fake followers
• Leveraged enhanced sampling techniques to maximize sampling within strict API constraints
• Identified and implemented classifier to detect fraud
• Delivered results on three week timescale

Irvine, CA

• Automated submission (bash), execution (bash), and statistical analysis (python) of batch free energy simulations to determine free energy profile of KvAP channel pore domain to evaluate energetic implications of.
• Modeled the Hv1 proton channel and performed molecular dynamics simulations to elucidate experimental results of functionally relevant alanine mutants
• Analyzed multi-microsecond simulations of Hv1 homology models with varying sequence alignments to optimize model and determine requirements for structural integrity in proton channel
• Calculated instantaneous structure factors of VSTx1 voltage sensor toxin with R and tcl scripting to assess validity of molecular dynamics simulations

Basel, Switzerland

• Thesis title: Mechanisms of the Coupling of Gating to Voltage Sensing in Transmembrane Proteins
• Modeled three dimensional Kv1.2/2.1 potassium ion channel in closed conformation from experimentally determined one dimensional distance measurements
• Predicted behavior of closed Kv1.2/2.1 potassium ion channel model and generated experimentally verifiable hypotheses from time resolved molecular dynamics simulations
• Contributed to development of NAMD interface for iPMF, a self-learning umbrella sampling software package for calculations of free energy landscapes
• Optimized expression protocol of KcsA for NMR studies of selectivity filter

Basel, Switzerland

• Thesis title: Computational Analysis of Secondary Structure Elements Relevant to the Gating of Voltage-Gated Potassium Ion Channels
• Wrote Python scripts for data mining of entire Protein Data Base for anomalously long 3-10 helices for statistical analysis of amino acid composition
• Performed molecular dynamics simulations to reveal the potential role of 310 helices in voltage-gated potassium ion channels

Basel, Switzerland

• National Champions, Switzerland Division Two: 2010
• Led 50-member multinational semi-professional football team to organization’s most successfulseason
• Coordinated the second-highest scoring offense in Swiss football history, averaging over 50points per game

Santa Monica, CA

• As the sole chemistry teacher at St. Monica’s, designed and implemented chemistry curriculum for 2nd, 3rd and 4th year students
• Received consistently positive evaluations from students
• Earned high praise from school administrators during formal observations for knowledge of subject matter, effective use of technology, and ability to motivate students, among others

Bachelor Of Science (B.S.), Biology, General

Doctor Of Philosophy (Phd), Biophysics

Master Of Science (M.Sc.), Molecular Biology

Education record