About Me

I am a PhD candidate in the joint Computer Science and BioInfomatic degree program at Clemson University and Medical University of South Carolina with a focus in Bioinfomatics, Data analysis, and Visualization. My interests lie in developing deep learning and machine learning algorithms to detect unique patterns in genomic and transcriptomic data./p>

Work Experience


Clemson University and Medical University of South Carolina (MUSC)

Worked with large scale RNASeq data including panCancer and panGtex to construct, normalize, and analyze gene and transcript expressions. Developed and implemented machine learning algorithms to extract gene correlation networks associated to brain cancer. Constructed one of the first gene regulatory networks associated to cerebellar tissues using machine learning. Developed and taught courses for Database Design and Management, and Introduction to Machine Learning.

Summer 2015, Summer 2016, Summer 2017

Lawrence Livermore National Laboratory

Developed visualization tools to identify performance improvement techniques for scientific application simulations. Identifying and visualizing spatial relationships between application and hardware for multivariate data. Utilized machine learning algorithms to explore correlation between application and hardware domain.

Summer 2013

Argonne National Laboratory

Developed a simulation of an integrated security system over a Light Weight File System written in C. Analyzed the effects of authorization and authentication servers on the timing effects to complete operations such as read/write.


Clemson University

Developing machine learning and visualization techniques to explore multivariate data. Developing visualization techniques and GUI to explore large scale calling context trees.



Technical editor pertaining to articles, papers, journal submissions in the field of Computer vision and pattern detection.


California Polytechnic, SLO

Investigating techniques for lip detection in low quality videos with subjects in moving vehicle. In-depth literature review of existing lip detection techniques and their limitations.

Technical Skills

Python (NumPy, SciPy, SciKit, Matplotlib), R

C, C++, Matlab, P-SPICE, Simulink, Processing

Visit, vtk, Paraview, QT, OpenGL

DirectX, MPI, OpenMP, OpenCV, Tableau

Ensembl, ToppGene, UCSC Genome Browser,

The Cancer Genome Atlas, Gtexportal

Research Projects

  • EdgeScaping: Mapping the Spatial Distribution of Pairwise Gene Expression Intensities.

    We address the concerns of restrictive linearity constraints in gene coexpression networks by implementing a deep learning algorithm to learn latent non-linear features.

    HeatTree: Exploring calling context trees.

    Developing techniques to reduce large scale calling context trees using structural similarity and pattern matching. Developing corresponding GUI using QT and OpenGL.

    Spatial RANSAC:Correlating multivariate scientific application and hardware domains

    Developed algorithm and GUI to explore correlation between multivariate scientific data and hardware counters using RANSAC sub-space clustering and mutual information techniques.

    Visualizing Fine-Grained Memory Accesses using Linked Software and Hardware Views.

    Developed framework and GUI to link inefficient memory access patterns from the source code to the underlying architecture using functional call graphs.

    Relating Memory Performance Data to Application Domain Mesh using an Integration API

    We have developed an API for Visit (a visualization and graphical analysis tool for viewing scientificdata) to integrate an external performance profiler or tool. This integration allows the user to leverage the visualization powers of Visit while mapping performance data onto its scientific domain.

    Overcoming Keyloggers and Screendumps

    Keyloggers (hardware or software) and screendumps of virtual keyboards by the local machine. To counter these attacks, we use DirectX 9 libraries on Windows or Linux operating systems. Our approach uses a remote server that communicates securely with the local process. The Direct X mode that we use executes in the GPU while being directly displayed on the screen. There is no direct communication between the operating system and the GPU storage, which allows us to communicate with the user securely even if the local machine is compromised. We present a simple prototype application of this approach, which supports web browsing.

    Domain-Name Generation Algorithm

    We implement two domain name generation algorithms using Probabilistic Context Free Grammar (PCFG) and Hidden Markov Model (HMM) to evade detection routines. We compare our DGAs with three popular detection routine Kullback-Leibler (KL) distance, Jaccard Index (JI) and Edit distance (ED) detection techniques.

    Security and Performance Evaluation of Security Protocols

    We evaluate 'Designed-In-Security' systems implemented for a distributed file system of exascale capacity. Designed-In-Security systems need to be evaluated for its capability to design, develop, and evolve high-assurance software, which is predictable and reliable while managing risk, cost, schedule, quality, and complexity.

    Exascale performance file system with integrated security system

    We simulate the authentication and authorization systems on a distributed file system to anyalyze the effect on performance with the addition of a security system. Variation of cache sizes and patterns of I/O operations significantly affected the system performance. We also implement and analyze an authentication and authorization prototype, which is a daemon based on Light Weight File System (LWFS), and simulated client and file servers to run performance tests to compare the simulation vs implementation performance difference.

    Face Detection and Lip localization in Moving Vehicles

    Developed an algorithm to detect lips (following face detection) on low quality images in a moving car environment. Evaluated Viola-Jones algorithm for test cases on lips and non-lips images in several color spaces.


    Benafsh Husain, F. Alex Feltus. EdgeScaping: Mapping the Spatial Distribution of Pairwise Gene Expression Intensities. In PLOS ONE, Accepted, July 2019.

    Yu Fu, Lu Yu, Benafsh Husain, Oluwakemi Hambolu, Ilker Ozcelik, Karan Sapra, Richard Brooks. Stealthy Domain Generation Algorithms in IEEE Transactions on Information Forensics and Security, vol. 12, no. 6, pp. 1430-1443, June 2017.

    Yu Fu, Benafsh Husain, and Richard R. Brooks. 2015. Analysis of Botnet Counter-Counter-Measures. In Proceedings of the 10th Annual Cyber and Information Security Research Conference (CISR '15). (Best short paper).

    Benafsh Husain, Alfredo Gimenez, Joshua A. Levine, Todd Gamblin, and Peer-Timo Bremer. 2015. Relating memory performance data to application domain data using an integration API. InProceedings of the 2nd Workshop on Visual Performance Analysis (VPA '15).

    Benafsh Husain, Alfredo Gimenez, Todd Gamblin, Peer-Timo Bremer, Joshua A. Levine. Visualizing Fine-Grained Memory Accesses using Linked Software and Hardware Views. IEEE Vis 2015, Poster.

    Alfredo Gimenez, Benafsh Husain, David Boehme, Todd Gamblin, Martin Schulz. Mitos: A simple Interface for Complex Hardware Sampling and Attribution. SC’2015, Poster

    Richard Brooks,Benafsh Husain, SeokBae Yun, and Juan Deng. 2013. Security and performance evaluation of security protocols. In Proceedings of the Eighth Annual Cyber Security and Information Intelligence Research Workshop (CSIIRW '13). ACM, New York, NY, USA.

    Karan Sapra, Benafsh Husain, Richard Brooks, Melissa Smith. Circumventing keyloggers and screendumps", in Proc. MALWARE '13, 2013, p. 103- 108.

    Benafsh Husain. Face Detection and Lip Localization. Master's thesis, Cal Poly, SLO, 2011.

    Xiaomin Jin, Sean Jobe, Simeon Trieu, Benafsh Husain Jason Flickinger. Mode pattern analysis of gallium nitride-based laser diodes. Proc. SPIE 7382, International Symposium on Photoelectronic Detection and Imaging 2009: Laser Sensing and Imaging (August 28, 2009).