Short Biography

Funny Me!</b>

My journey into the wonderful realm of computing began when I joined Allameh Helli School, affiliated with the National Organization for Development of Exceptional Talents (NODET). I quickly learned the foundations of algorithmic design and programming in C/C++. In this process, I implemented an IPX-based framework for cluster/high-performance computing, which won the prestigious National Kharazmi Award in the 4th annual Kharazmi Youth Festival, 2002. Later, I improved over my initial design and acquired the intellectual property (IP) for it in 2004. Parallel to my research on parallel/distributed computing, I also gained interest in robotics. Together with three other students, we initiated the first student group ever to participate Robocup competition in 2003 in Padova, Italy.

Following the same path, I chose to join the University of Tehran in 2004 to continue my education in computer science. During first few years, I have been working on the similar set of problems in robotics and parallel computing. These projects gave me a solid background in machine learning/AI, as well as a mastery in designing and implementing efficient parallel/sequential algorithms. During last year of my Bachelor's study, I joined the Center of Excellence in Bioinformatics. As part of the center, I got introduced to the field of computational biology with a fascinating problem in systems biology: to identify reoccurring building blocks in complex networks that are significantly overrepresented compared to an ensemble of random networks, which are also known as network motifs. This project resulted in a well-cited publication and highly used software package called Kavosh, which was later ported to the commonly used Cytoscape framework. During this phase of my life, I learned about another problem in network biology: how to align interaction networks of distant species, also referred to as network alignment. I started my research on this topic prior to starting graduate school when I learned about the work of my Ph.D. advisor and one of his former student, Mehmet Koyuturk. Being intrigued by the intellectual merits of their work, I decided to join Purdue University.

As a graduate student at Purdue University, I work under the supervision of Ananth Grama and Wojciech Szpankowski and collaborate closely with Shankar Subramaniam (UCSD), Mehmet Koyuturk (Case Western Reserve), and Andrea Goldsmith (Stanford University). During this period, I had the opportunity to work on a broad range of problems spanning different areas of computational biology. In the realm of network/systems biology, I have extended network alignment methods in various directions both in terms of efficiency and effectiveness, the latest of which is a method called Triangular Alignment (TAME). This method is a tensor-based approach to combine network motifs with network alignment. The premise is that if we score aligned motifs, such as triangle motif, instead of merely aligned edges, we can significantly enhance the quality of alignments in terms of both conserved topology and, more importantly, their biological underpinnings. On the application side, I have applied complex network analysis techniques to (i) study aging-related pathways in yeast, and in a follow-up project, (ii) assess the conservation of yeast pathways in tissue-specific human.

While working on these problems, I developed a curiosity to understand how these analyses extend if you zoom into a specific cell type/tissue context. This curiosity was triggered by a fundamental question: how a single-cell organism, such as yeast, can be used as a model to understand tissue-specific pathways in humans? How much of this biology is conserved? As an example, yeast is the dominant model organism for aging research. However, I was not sure how well our findings in yeast could shed light on the onset and progression of age-related disorders, such as neurodegenerative diseases and cancers. Motivated by this problem, I started developing methods tailored towards extending traditional techniques along spatiotemporal context of cells. Since then, I have developed a variety of tools to (i) identify cell types from single-cell transcriptional profiles, (ii) deconvolve transcriptional profile of complex tissues, and (iii) constructing tissue-specific interactomes.

My future direction is mainly to gain more hands-on experience with experimental techniques and clinical applications. I am keenly interested in learning more about the design of different technologies to assay -omics datasets. Following my latest passion in tissue/cell type-specific analysis, I am interested in understanding the process of intercellular signaling pathways. Unfortunately, there is no commonly used high-throughput experimental procedure for accurately measuring intercellular events at high spatiotemporal resolution. I believe that there are promising avenues for exploring this space using genome barcoding, followed by protein fusion, coupled with mass spectrometry. This body of knowledge is crucial for understanding tumor-immune interactions, as well as for personalization and enhancing the efficacy of immunotherapies. They also provide a mechanistic understanding of resistance to targeted therapies in different subclasses of tumor cells, in different patients, and different stages of cancer. This, in turn, suggests alternative targets for combination therapy and to combat resistance.