Eli Draizen

PhD Candidate in Phil Bourne's lab @ University of Virginia

About Me

Hi, I am a PhD Candidate at the University of Virginia in the labs of Phil Bourne and Cameron Mura. I am interested in structural bioinformatics and machine learning applied to protein evolution and host-pathogen interactions. I was previosly an organizer for the ISCB Region Student Group in DC as the Student Council Symposiums at the main ISCB conference.

You can download my full CV here.


University of Virginia

Ph.D Biomedical Engineering

Expected 2021

Under the guidance of Phil Bourne and Cameron Mura, I am applying 3D Convolutional Neural Networks to predict protein-protein interactions (PPI) using protein structure information. This work will be used to identify drug binding sites to disrupt PPI and proteins involved in host-pathogen PPI. This work is also done in collaboration with Anna Panchenko and Alex Goncearenco at the NCBI.

I am also working on developing deep learning methods such as Variational Autoencoders for protein structure classification. We hope to use this to find new relationsips between CATH homologous superfamilies.

UC Santa Cruz

B.S Bioinformatics

June 2014

Thesis: Correcting Frameshift Mutations in Transcriptomics Data

Adivsors: Mark Blaxter (Edinburgh) and Dietlind Gerloff



1. Daniele Parisi *, Gabriel J. Olguín-Orellana *, Eli J. Draizen *, …, and R. Gonzalo Parra. Nurturing tomorrow’s leaders: The iscb student council symposia in 2018. F1000Research, 2019. pdf | abstract

2. Thomas A. Hopf, …, Eli J. Draizen, …, Chris Sander, and Debora S. Marks. The EVcouplings Python framework for coevolutionary sequence analysis. Bioinformatics, 2018. pdf | abstract

3. Cameron Mura, Eli J. Draizen, and Philip E. Bourne. Structural biology meets datascience – does anything change? Current Opinion in Structural Biology, 2018. pdf | abstract

4. Nigel W. Moriarty, Eli J. Draizen, and Paul D. Adams. A restraints editor for generationand customisation of geometry restraints. Acta Cryst. D, 73(2):123–130, Feb 2017. pdf | abstract

5. Lei Xie, Eli J. Draizen, and Philip E. Bourne. Harnessing big data for systems pharmacology. Annu Rev Pharmacol Toxicol., 57(1):245–262, 2017. pdf | abstract

6. Eli J. Draizen, …, David Landsman, and Anna R. Panchenko. HistoneDB 2.0: a histone database with variants – an integrated resource to explore histones and their variants. Database, 2016:baw014, 2016. pdf | abstract


1. Menuka Jaiswal, …, Eli J. Draizen, …, Cameron Mura, and Philip E. Bourne. Deep learning of protein structural classes: Any evidence for an ’urfold’?. arXiv:2005.08443, 2020. pdf | abstract

2. Sean Mullane, …, Eli J. Draizen, …, Cameron Mura, and Philip E. Bourne. Machine learning for classification of protein helix capping motifs. arXiv:1905.00455, 2019. pdf | abstract

In Preparation

Nigel W. Moriarty, Eli J. Draizen, and Paul D. Adams. CarboLoad: Modelling, validation, and prediction of carbohydrates in phenix.

Eli J. Draizen, D. Steve Hall, Felecia D. Kemp, Jonathan Magasin, and Dietlind L. Gerloff. Plasmodium 6-Cys model database.

Eli J. Draizen, Edward Y. Liaw, Jonathan Magasin, D. Steve Hall, Felecia D. Kemp, and Dietlind L. Gerloff. Tracking the evolutionary link between distant Apicomplexan surfaceprotein families by their distinct disulfide patterns.


Wrote HistoneDB 2.0, a Django webserver to store and classify histone sequences by variant using an ensemble of hidden Markov Models. Work done in Anna Panchenko’s lab at the NCBI

A webserver to study the structure and evolution evolution of the ‘6-Cys’ domain containing proteins proteins from Plasmodium falicparum, the cuasitive agent of malaria. Initial version for Plasmodium created in Dietlind Gerloff’s lab at UC Santa Cruz, but in the future it will include all Apicomplexan parasites.

Developed CarboLoad, a program to validate and model glycoproteins by ligand fitting, and Carbohydrate Builder, a 2D carbohydrate builder and visualizer supporting GlycoCT. Work done in Paul Adams lab at Lawrence Berkeley National Labs

Developed HSP-Tiler, a python script to mend frameshift mutations in transcriptomics data using BLAST and HMMs for senior thesis at UCSC Work done in Mark Blaxter’s lab at The University of Edinbugh

Perform model mixtures of different hidden Markov Model packeges (HMMER and UCSC-SAM), saving the the results into a pile up.

Fork me on GitHub