About Me


Welcome to my humble online abode! My name is Levon Demirdjian and I’m a freshly minted PhD from the Department of Statistics at UCLA. My guiding philosophy in work, as in life, is simple:

Improve human health and well-being by harnessing the powers of big data.

As a professional statistician/data scientist, my research has focused on developing models for extracting knowledge from the vast amounts of biological and genetic data being generated by next-generation sequencing methods like RNA-Seq. I am convinced that such models will not only generate important scientific breakthroughs but will ultimately contribute to a more thorough understanding of human health and provide potential avenues for combating illness and disease.

I completed my PhD under the guidance of two extraordinary advisors; Dr. Ying Nian Wu in the Department of Statistics, and Dr. Yi Xing in the Department of Microbiology, Immunology & Molecular Genetics.

Update: 07/12/2018

I am incredibly excited to be joining Dr. Yi Xing in August at The Children’s Hospital of Philadelphia as a postdoctoral fellow in the newly founded Center for Computational and Genomic Medicine! Read the press release for our new center here.

Decoding the human genome

We live in an age of unprecedented scientific and technological breakthroughs which are largely attributable to so-called “big data.” However, the generation of data, i.e. information, is not the same thing as the generation of knowledge. Nowhere is this disparity between information and knowledge more apparent than in the realm of human genomics.

Our genome is composed of nearly 3 billion base pairs, and keep in mind that this information is located in nearly every cell of our bodies. To put that staggering number into perspective, the Wellcome Trust in London created a physical library of books containing the entire DNA sequence of a human reference genome. According to the Wellcome Collection, “The 3.4 billion units of DNA code translate into over a hundred volumes, each a thousand pages long, in type so small that it is barely legible.”

Printout of a human reference genome displayed in the Wellcome Collection in London. Image Credit: Russ London at English Wikipedia, CC BY-SA 3.0
One of the 1000 pages from one of the nearly 100 books in the Wellcome Collection.

Now, if information and knowledge really were the same thing, we should be able to look at the raw genetic data pictured above and learn something meaningful about this particular stretch of DNA. Cleary this is not the case (unless you have some superhuman capabilities).

How do we begin to make sense of this deluge of data?

Enter bioinformatics, a discipline which leverages computer science, biology, mathematics, statistics, and machine learning in order to extract meaning from biological data. Modern advances in deep learning algorithms and computing power have allowed bioinformatics researchers to yield insights into the genome and transcriptome which would have been unattainable even 10 years ago.

Deep learning framework for analyzing alternative splicing using genomic sequences and RNA-seq data (Park et al 2018).

Not only have these modern technological advances shed light on the complex biological processes contributing to diseases like cancer, but they have also helped pave the way for novel approaches like immunotherapy for combating these diseases.

My Name

Levon (in Armenian Լեւոն, pronounced “Lev” as in “level” and “on” as in “bone”) is a typical Armenian name given to boys. The name is mostly attributable to a line of Armenian kings that ruled Cilicia starting in the late 12th century A.D. with the reign of Levon I; under Levon I’s rule, Cilician Armenia was established as a powerful and unified Christian state. The line ended in the late 14th century upon the surrender of Levon V to invading Ottoman forces. At the height of its power, the kingdom of Armenia was nearly 10 times as large as the current Republic of Armenia.

Demirdjian (in Armenian Տէմիրճեան, pronounced deh-MEERJ-yawn) roughly translates into “son of blacksmith”, so my ancestors must have been metalworkers at some point in time. Armenian last names almost always end in either “ian” or “yan”.

And in case you were wondering, Kardashian (not to be confused with Cardassian) literally translates into “stone sharpener”, so our beloved celebrity family descends from a long line of stonemasons. Consider this my public service announcement for the year.

I Am Currently Reading…

  • Winter is Coming (Garry Kasparov)
  • The Gambler (William C Rempel)
  • Anything and everything in Significance Magazine