Over the past decade, high-throughput technologies have transformed biomedical research, generating substantial amounts of data (big data). Deciphering meaningful biological insights from this data poses a significant challenge . My research focuses on the development of algorithms and analytical approaches for genomic, proteomic, and glycoproteomic data integration and analysis. I also have extensive experience in creating web-based genomic and proteomic databases. Recently, I established an aqueous humor proteomics database. Using a similar platform, I plan to create a reference database for tear fluid proteins and miRNAs, providing a freely accessible resource for vision researchers.

My research focus is to develop computational methods for high throughput data analyses. Currently, I am working on analytical methods for the large-scale measurements that are part of several Type-1 Diabetes and Cancer research projects. Specifically, I am analyzing data from many high-throughput technologies, including Luminex, gene expression microarrays, SNP microarrays and mass spectrometry-based proteomics and metabolomics. I am also working towards the integration of these genetic, genomic, and proteomic data to reveal new biological insights.

• To develop algorithms and analytic approaches for genomic data analysis and integration. • Solve the challenges in genomic data analysis, and implement these solutions in usable software tools.• Working with an interdisciplinary team, carry out bioinformatics analyses, and pursue research projects in collaboration with other biologists.• To integrate microRNA expression and mRNA expression data. • Independently utilize web resources (such as Entrez, RefSeq, GEO, BLAST), visualization tools, genome assemblies, and available analysis software to analyze genomic data.

mRNA-microRNA modules associated with leptin mediated weight loss: I developed a general framework that integrates genomic data to identify key pathways and mRNA- microRNA modules associated with disease or given phenotype. As a proof of principle, applied these methods to reveal gene regulatory networks correlated with leptin mediated weight loss. Microarray data analyses: I have worked closely with biologists to facilitate the microarray data analyses and interpretation. I have performed a variety of statistical analyses on gene expression experiments including normalization, differential expression analyses (SAM, LIMMA), cluster analysis, gene ontology analyses, and pathway analyses by utilizing self- developed and publicly available resources. HPCluster: I developed a software tool called HPCluster to cluster extremely large gene expression datasets using a two-stage algorithm. HPCluster uses a single scan to pre- partition the data before subsequently clustering using a simple k-means algorithm. The data reduction step used in the first stage of algorithm reduces the memory requirements allowing us to cluster extremely large datasets without failure. ParaSAM: To overcome the memory limitations, I developed a parallelized version of the Significance Analysis of Microarrays (SAM) algorithm called ParaSAM. The parallel nature of the application comes from the use of web services to perform the permutations. ParaSAM is not only faster than the serial versions, but can analyze extremely large datasets that cannot be performed using the existing implementations.