The focus of our research is to accelerate the understanding of basic mechanisms involved disease, speciation and the interplay of the genome, transcriptome, epigenome, proteome and the rest of the “–omes”. This is done through the judicious generation and analysis of local data that is then analyzed in context with the larger body of knowledge found within public databases that contain genomic data across species, expression (RNA and protein) data and other meta data associated with sample annotation, especially clinical samples.

Our work spans the range from de novo genomic sequencing, to re-sequencing, deep sequencing, genotyping, epi-genotyping, bulk cytogenetic and histologic inspection, phenotype measurements. The overall intent is to expand our knowledge and understanding, to improve medicine and to improve the conditions of humans and other species on the planet earth. The developments in our lab are applied and validated by investigating cancer, cardiac disease (the number 1 and number 2 killers of Americans) and other diseases such as arthritis, MS, epilepsy, indeed any disease to which our platform technologies can be applied. Frequently our work is done in collaboration with various esteemed researchers across the world. Of particular note is our work with the cancer center and NIH SPORE in lung cancer at UTSW, the NHLBI Proteomics Center and the Western Regional Center of Excellence in Biodefense and Emerging Infectious Diseases (WRCE).

Of particular note, our work involves the prediction or identification of genetic variation hot spots, especially those that are likely to result in a measurable phenotype or a disease. In this area we analyze microsatellite polymorphisms data as well as single nucleotide polymorphisms (SNPs). These known or predicted hot-spots for polymorphism in the genome can play important mechanistic roles in genetics via protein sequence or splicing alteration, expression modulation via promoter alterations, epi-genetic alterations such a adherent methylation, and changes in protein-DNA binding properties.

Areas of Emphasis in Genomics/Proteomics

Universal Bio-Signature Detection Array Global directed search for polymorphisms, Gene Expression Analysis, Mass Spec Analysis Pipeline, Cancer Research, Cardiac Disease Research, Light Biology, Patterning device and associated chemistry for tissue engineering

Global directed search for polymorphisms - Our efforts are both computational and experimental, and span both the identification of single nucleotide polymorphisms (SNPs) and microsatellites (repetitive DNA sequences). Our initial work was the development of computational approaches to find repeats in genomic sequence, especially repeats or microsatellites that are likely to be polymorphic. That enabled us to develop a catalog and to pursue several applications areas, especially in the development of new cancer biomarkers and possible causative polymorphic repeats, but also in the area of development, where we focused our effort in those polymorphisms that are likely to alter morphology in a quantitative way. Both of these areas resulted in numerous publications, yielding markers for lung, colon and breast cancer as well as even finding the repeat responsible for setting the length of a dog’s nose. These computer predictions are available in our web-based database/tool, EREMORPH.

We feel that microsatellite polymorphism and its role in disease, development and even evolution is understudied. Most groups study SNPs because they have been implicated in a number of different diseases, and because they are abundant and easy to measure. Microsatellites are more challenging, and to date there has been no way of measuring them en masse. We have developed a global microsatellite content array, which measures the total integrated amount of a given microsatellite repetitive motif in a genome that is hybridized to the array. We have applied that array to the areas of evolution, development, cancer and even forensics. Those studies are underway, and are yielding new ways to classify cancer, identify new cancer biomarkers and even may help determine the genomic variations that are responsible for some of the differences among species. This work is underway and will hopefully be reported in the literature soon.

Lastly, we have also developed computational techniques to predict the potential impact of particular SNP variations within coding regions of genes, i.e. those SNPs most likely to cause a phenotype of a disease. This approach was recently reported in a manuscript, and the results are also available for every coding base in every gene in the database/tool, EREMORPH. Briefly, we compared those SNP base positions that were known to cause disease (using the database HGMD) and those that are less likely to cause disease, all those SNPs in dbSNP (that are also not in HGMD), and together these two training sets were subjected to a Support Vector Machine, allowing us to then use the resulting trained network to assign a probability for impact to all coding base positions in a gene.

Gene Expression Analysis - Gene Expression Analysis - We have been developing new protocols for improved standardization in the analysis microarrays. The group, as part of our work as the Bioinformatics Core for the Western Regional Center of Excellence in Biodefense and Emerging Infectious Diseases (WRCE), we offer turnkey expression analysis and interpretation. This work includes consultation with researchers in the design phase of an experiment, followed by analysis and interpretation of the data, supplying manuscript ready text, tables and figures. Another unique component of this work is the use of multiple microarray analysis software algorithms to ensure that all genes found to be differential are truly robust, i.e. are reproducibly and reliably so, independent of the particular algorithm used, for we pass all data through many, typically 5 algorithms, keeping only those genes that survive all tests. It should be noted that we also have applied this analysis process to a number of collaborator’s data, in the areas of cancer and cardiac disease research.

In addition to the traditional gene list data analysis techniques, we have developed and regularly apply a text mining engine, IRIDESCENT, to the plurality of genes that emerge from expression or other experiments. This has allowed us to find new ways to classify cancer and to find hidden meaning in the set of genes.

Lastly, we have also developed a number of ancillary techniques for improving the experimental execution of expression studies. This has included the development of genomic labeling (where DNA is used as the standard) and with RNA pools developed by Stratagene, as well as developing internal standards for microarrays that can separate RNA contamination from CGH experiments, for example.

Mass Spec Proteomics Analysis Pipeline – The analysis of the tremendous amount of data that emerges from a high resolution mass spec device is challenging and limited by the process used to uniformally prepare samples for analysis as well as the techniques for data quality control evaluation, data trimming and even the algorithms used in the final identification of differential ‘peaks’ in spectra that may be biomarkers. As such, we developed algorithms to address these issues, have published them and offer this analysis service to collaborators. Our substantial computer capability, specifically, large clusters of 64-bit machines are necessary for the analysis of these very large, terabyte sized data sets. We are currently working to develop standards and to understand natural variation in sera spectra in humans by studying that variation longitudinally over long and short periods of time. We hope that this work will contribute to reducing the number of false positive peaks found to be differential as a consequence of this natural variation.

Cancer Research - In close collaboration with John Minna, M.D., our group has been developing and then applying our genomic sequence analysis tools, especially Panorama and Pompous, to study lung cancer. With John, we have analyzed in depth human chromosome 3p for tumor suppressor genes and genetic variations that contribute to cancer progression. This research is expanding to incorporate identification of methylation and regulatory control elements of cancer genes in software and then laboratory analysis on expression microarrays and re-sequencing (DOC) arrays. An example: With John Minna, the following demonstrated for one of our predicted simple sequence repeats that it was polymorphic and was informative in that it could differentiate different patients and the state of the lung cancer tumor (LOH was observed in some patients tumor DNA compared to their normal cell DNA). We also constructed and maintain the database of all experimental data emerging from the NIH/NCI Lung SPORE called SporeBase. We also have developed a number of other unique tools and databases for the SPORE, as a Core in that large program project grant.

Cardiac Disease Research - We are working with several groups at UTSW to understand the genetics of heart disease. To identify the genes that contribute to heart disease and risk, we are identifying candidate genes using experts, microarray analysis of mouse models, and our computer codes. At least 600 genes identified were completely re-sequenced in substantial sets of cohorts to identify polymorphisms, providing the basis of subsequent mass spec-based assays will be developed by Sequenom and SNP analysis using Perlegen microarrays to then genotype 3,500 patients collected by the Reynold's Center.

Light Biology - In the area we now call light biology, we have been investigating new biomedical applications of light projection and detection. In the area of light projection using Texas Instruments Digital Light Processing chips (DLP), this has included the development of the DOC system for chip-based re-sequencing and expression analysis, modulation of the immune system using specialized light sources, tissue engineering. This formed the basis for a company, later sold to Nimblegen, Inc (now Roche). We have now constructed a peptide/peptoid custom microarray synthesis technology based on our initial work manufacturing oligonucleotide microarrays. In the area of light detection, we have focused on developing cytogenetics, re-sequencing and expression methods that use hyperspectral imaging to enable more accurate determination of labeled probes and a higher depth of probe multiplexing.

Patterning device and associated chemistry for tissue engineering-We have devised and built several patterning systems with resolution down to 0.5 microns, fully time dependent control of the pattern and control over the light source for illumination (visible, UV, IR, pulsed laser). This enables us and our collaborators to precisely control the deposition of chemicals/drugs and also control the activation and deactivation of various processes at an organelle resolution. When coupled with various fluorescent labeled protein biomarkers, the readout of the cellular response is achieved at with high spatial and time resolution.