Resume
Saraswathi Sundararajan, PhD
Cell: 1-515-***-****
Email: **.*****.**********@*****.***
Citizenship: Singapore;
Permanent Resident of USA (Green Card holder)
Current status:
Postdoctoral Research Associate, since Aug. 2011 to present. The Research Institute at
Nationwide Children’s Hospital,
Battelle Center for Mathematical Medicine, Columbus, OH 43205, USA
Education
1. PhD: Iowa State University, Bioinformatics and Computational Biology. Thesis: Protein
Secondary Structure Prediction using Neural Networks and Particle Swarm Optimization
2. M. Sc.: Nanyang Technological University, Singapore, Bioinformatics,
Thesis: Cancer Classification of Microarray data, using Machine Learning techniques
3. M.S.: Old Dominion University, Norfolk, VA, USA Computer Science
4. E-Commerce Engineering and Java Certification, UC Santa Cruz, USA.
5. BA (Mathematics and Economics), University of Delhi, India
6. CCNA certified and completed courses in CCNP and other Windows administration courses.
7. Additional courses in finance and management (ACCA, Singapore), Biology and Statistics
at Masters level.
Current Research: using machine learning in computational biology
Methods: Genetic Algorithms for feature selection; Neural networks and Particle Swarm
Optimization for classification using MATLAB, WEKA and other bioinformatics tools.
• Collaborator: Dr. Peter Houghton: Developing methodologies for molecular
therapy response classification of childhood cancers in collaboration with Prof. Peter
Houghton. Machine learning and advanced statistical algorithms are used to identify molecular
biomarkers that predict efficacy within the context of PPTP childhood cancer pop ulations. An
Integer-Coded Genetic Algorithm (ICGA) is used for feature selection, data reduction and
identification of biomarkers. A Neural Network based Extreme Learning Machine (ELM)
algorithm in combination with Particle Swarm Optimization (PSO) is used to classify and
differentiate between different classes of data. We are interested in discovering possible
mechanisms of resistance or sensitivity that could in turn be targets of a synergistic therapy or
candidate genes for molecular perturbation studies. We have de monstrated that a relatively small
gene-expression signature derived from in vitro sensitivity data (using Genetic Algorithm and
neural network) is able to accurately predict the overall in vivo response at a of 80-85%.
• Collaborator : Dr. William Smoyer and Dr. Richard Ransom : (one R01 grant
application) Paired blood samples (Rna_seq data) from patients with Nephrotic Syndro me are
used for IPA analysis, classification and prediction of steroid resistant vs. steroid sensitive
patients. Although the samples were very few, we were able to reduce the microarray data set
using Genetic algorithms and clustering (collaborative study) to reduce the features and perform
machine learning analysis using neural network. Various statistical analysis were also performed
on the data.
• Three projects in collaboration with Dr. Chotani Maqsood: (one R01 grant
application)
The first project relates to the examination of signature microRNAs and proteins in
o
human vascular smooth muscle cells utilizing chip/PCR analyses and 2D-proteomics
(fluorescence difference gel electrophoresis), respectively. This approach will potentially lead to
identification of physiologically relevant "vasculo-protective" players (or markers), conferring
"stress-tolerance" to cells during cooling or vascular injury. Our preliminary studies using our
machine learning and feature selection techniques have illustrated that there are marked
differences in data relating to male and female mice cells which was a new discovery (our
contribution) in this research. We are in the process of identifying bio-markers related to
this study. We are using IPA analysis to study the genes selected using our methods.
The second project relates to Role of the Ras-related small GTPase Rap1A in the
o
cardiovasculature Rap1 acts as a molecular switch, coupling extracellular stimulation to
intracellular signaling through the second messenger cyclic AMP. The key objective of this
project is to examine the role of this protein in cell attachment and cell survival in
vascular smooth muscle cells, and in cardiac myocyte function. We are processing the
miRNA data for this study and are working on identifying markers using our algorithms as
described above.
The third project relates to the investigation of the alpha-2A and 2C adrenergic
o
receptors. Several bioinformatics tools were used to analyze these proteins for predicting
secondary structure, disordered regions, binding regions and relative solvent accessibility.
I-Tasser was used to build protein models and nor mal mode analysis was used to for dynamic
analysis.
• Collaborative study with Dr. Octavio Ramilo and Dr. Maria Mejas: Microarray data on
Children’s infectious diseases were provided for several diseases such as flu, RSV, RV and
controls. We have processed these data using Neural networks and GA algorithms to identify
genes which might be responsible for these children’s diseases and have performed IPA
analysis for these genes. This study was done last year in 2012.
• Collaboration with Dr. Lucchesi and Dr. Vidu Garg and Dr. Loren Wold. Discussion
of projects are going on.
• Collaboration with Dr. Bakaletz on modeling of biofilm. On going talks.
• Collaboration with Dr. Petril of OSU and Dr. Bartlett of BCMM, NCH on
ADHD studies : On going talks.
• Collaboration with the perinatal group and Dr. Sarah Kleims : on-going talks.
Continuing Research from my PhD Studies
• Protein secondary structure prediction, Solvent Accessibility, Phosphorylation etc.
• Investigating use of protein physical and chemical properties in secondary structure
prediction (using data from AAindex database)
• Investigating use of Position Specific Residue Preferences (PSRP) of amino acids in
structure prediction
Professional Experience 2007 – 1995 (2006 to 2011, Bioinformatics MS and PhD)
• Visiting Researcher: Bioinforamtics Research Center: Machine Learning research,
School of Computer Engineering, Nanyang Technological University, Singapore. (July 2006 to
July 2007).
• Bioinformatics Research Center: Amrita University: I worked on Machine learning
(SVM) techniques, and parallel processing using grid computing and MPI for
bioinformatics applications. (Jan 2005– June. 2006)
• Business Consultant: at Amritha Viswa Vidya Peetham, Coimbatore, India: Project
Lead for development of financial modules and Enterprise Resource Planning Software
developed for Konkan Railway System. Later joined as Visiting Researcher at the Center for
Networking and Excellence, Amrita University, Coimbatore, India. (Jan 2004– Dec. 2004)
• Software Engineer and Technical writer, ValueLabs, Hyderabad, India : Test
development and quality assurance for mobile portal applications. for mobile portal
applications such as interactive TV and multimedia messaging. (June 2002 Web page and web
page application development – Dec. 2003)
• Software Engineer/Consultant, Techsoft Consultants, India and CA, USA 2001 – 2007
(Freelance Consultant, doing work undertaken as contract. Developed and administered
website for a major shipping company, Air7Seas. Developed test tracking application for a
multimedia software vendor, Margi Systems. Computer security consultant for VPN
Dynamics. Familiar with web technology, Microsoft Office packa ges, E-commerce engineering
and database concepts, Scripting and MySQL. Working knowledge of Unix and expert
knowledge in Windows environment. Knowledge of the following computer languages: Ada, C,
C++, Visual BASIC, Java, MATLAB, HTML, Perl and Python, which were used at various times
for software development and research activities. (Jan 2001 to June 2007)
• Visiting Researcher and Tutor: School of Computer Engineering, Nanyang
Technological University and National University of Singapore - Developed Finite Element
analysis modeling software package in C++. The ultimate objective of the software was to
simulate material behavior under extreme conditions. I implemented code migration of a
geometrical Finite Element Analysis package from ADA to C. Tutoring Business and
Accounting students in Information Technology (Jan. 1995 –2000 Dec.)
References available on request