Emmanuel R. Yera

Previous Research

GenomExplorer: genomic cancer data visualization
Tumor genomes can be highly rearranged and, thus may not be co-linear with host genome. Recurrent genome rearrangements involve genes that mediate a wide range of growth and signaling pathways that are increasingly targeted by anti-tumor therapeutics. Current technologies for studying tumor genome structure are not capable of elucidating the structural organization of tumor genomes at high resolution, or of relating it to the underlying host sequence. Consequently, the role of translocations and inversions in solid tumors is poorly understood. Indeed, even the structural organization of amplicons remains largely enigmatic. End Sequence Profiling (ESP) is a sequence-based method for directly determining the structural organization of tumor genomes, and for cloning all types of structural rearrangements en masse. In addition, ESP can be carried out on tumor transcriptomes for large-scale identification of fusion transcripts. We have demonstrated this by analyzing full length enriched and normalized cDNA libraries from MCF7 (breast cancer), LnCaP (prostate cancer), and a primary brain tumor. Multiple tumor-specific transcripts were identified and analyzed. The sheer amount and diversity of the data generated by ESP necessitate development of efficient and effective way to visualize and compare ESP genomic and transcriptome data. The purpose of this project, GenomExplorer, is to provide a mechanism for visualizing ESP data by means of a web based application leveraging Asynchronous Javascript and XML (AJAX) technology. We expect this tool to allow us to visualize ESP data from multiple genomes in order to facilitate the detection of common rearrangements which can later be validated in the lab.

Machine learning techniques for array comparative genomic hybridization (aCGH) cancer classification

Since the arrival of microarray technology there has been an explosion of gene expression profiling data. In spotted microarrays the probes which are placed on the array correspond to oligonucleotides, cDNA or small fragments of PCR products which correspond to mRNAs. The array is then hybridized with cDNA of two samples generally normal versus diseased, each sample is colored red or green. The intensities of the spots can then be measured to see if there has been gain or loss for a particular spot. Given a normal tissue and a tumor tissue microarray experiments can be executed to find out how the gene expression profiles differ. These types of experiments can lead to many types of discoveries such as gene modification in the tumor thus allowing the tumor to replicate faster than normal.

Given microarray experiments is there a way to do classification of different types of cancer? A lot of work has been done trying to answer this question, the quick answer is yes very accurately > 90 percent but this is not general for all cases and is very cancer specific. There are other types experiments which can be done to analyze tumors such as array comparative genomic hybridization (aCGH) which detects copy number changes. aCGH is similar to spotted microarrays except that instead of being interested in mRNA you are interested in the DNA. The purpose of this project is to perform feature selection on aCGH data to find optimal subsets which can accurately classify progressed from not progressed cancer.

Molecular Query Retrieval
Discerning the similarity between two molecules is a challenging problem in molecular biology as well as in drug discovery. Since similar molecules often have similar biological behaviors and chemical properties, molecular similarity is a key notion in biochemical investigations targeted at understanding existing molecules as well as guiding the synthesis of new molecules. Additionally, molecular similarity plays a key-role in structure-based query-retrieval in molecular databases as well as in computational investigations seeking to understand the structural space. Our experimental results hope to validate the properties and efficacy of the similarity formulation and underline its applicability for searching rapidly expanding structural repositories.

Publications
E. Velasquez, E. R. Yera, and R. Singh, “Determining Molecular Similarity for Drug Discovery using the Wavelet Riemannian Metric”, IEEE Symposium on Bioinformatics and Bioengineering (BIBE), pp. 261-268, 2006. pdf

R. Singh, E. Velasquez, P. Vijayant and E. R. Yera, “FreeFlowDB: Storage, Querying and Interacting with Structure-Activity Information for High-Throughput Drug Discovery”, IEEE International Symposium on Computer-Based Medical Systems (CBMS), 2006,75-80. pdf

Graph Theory
Description coming soon.
Publications
E. R. Yera, L. Wang and S.M. Lee, “On Super Edge-graceful Eulerian Graphs”, Congressum Numerantitum 174 (2005), 83-96.

E. R. Yera, W. Ling, E. Chen and S.M. Lee, “On super edge graceful (p,p+1) graphs”, Congressus Numerantium 171 (2004), 51-65.

E. R. Yera, W. Ling and S.M. Lee, “On super edge graceful unicyclic graphs”. Journal of Combinatorial Mathematics and Combinatorial Computing (2004).