Business analytics using Large Language ModelsClinical trial biomarker analysis

* Developed a novel computational platform to discover tumor microenvironment targets* Analyzed clinical trial biomarker data and prepared conference presentations* no direct reports

Scope and drive innovative projects with technology leaders from MIT, Broad Institute, UCSF:* CODEX protein and CITE-seq single cell analysis of immunological skin diseases* high sensitivity single-cell RNA-seq and spatial transcriptomics of skin* machine learning to discover immunological disease-target associations* predicting dermatological drug response from electronic medical recordsGuide company bioinformatic and AI activitiesno direct reports

* Developed machine learning methods to optimize antibody activity using phage display data including deep learning recurrent and convolutional network models based on amino-acid sequences* Developed digital biomarkers for Parkinson disease and chronic fatigue syndrome from accelerometer data* Designed and analyzed DNA-and RNA-SEQ studies* Built deep-learning hyperparameter optimization architecture on Azure Batch

• Developed computational pipeline to quantify mutation load in NGS mitochondrial genomes• Built a NGS genomic data analytical solution on Google Cloud including BigQuery storage and Tableau end-user frontend• Developed predictive model of ALS progression from patient voice data using deep neural network algorithms• Found biochemical markers of ALS progression from the PRO-ACT clinical trial database• Developed a computational pipeline for analyzing neuron cell image drug screens including machine-learning object recognition• Built a data warehouse, analytical and visualization tools integrating microarray and RNA-SEQ expression, biological pathways, ontology and protein interactions in order to discover pathways regulated in ALS• Administer the company Linux and MySQL bioinformatics servers

Performed analysis of expression and DNA copy (Agilent CGH platform) data.

• Performs microarray/QRT-PCR experiment design, data analysis and interpretation to derive markers of disease and treatment efficacy for respiratory diseases. • Built respiratory disease related biological pathways using text-mining technologies.• Developed a Spotfire/R module for interpreting high-throughput genomic experiment results using biological pathway information.• Developed statistical tools for comparing QRT-PCR quantification and normalization methods and identifying best normalization genes.• Developed Visual Basic Excel plug-ins to render and annotate expression analysis tables.

• Developed prototypes and directed developments of:- system integrating multiple algorithms and databases for providing comprehensive automatic annotation of genomic targets.- text-mining tools for extracting information from Medline and Derwent-GENESEQ abstracts.- protein cell localization prediction system based on sequence motif composition.- family classification, phylogenetic analysis, ortholog identification system.- computational identification of novel splicing variants.• Co-directed delivery of 900 gene knowledge packages to pharmaceutical partner, Bayer AG.• Built comprehensive collections along the sub-family classifications for human protein kinases, ion-channels, nuclear hormone receptors, G-protein coupled receptors.• Provided technical leadership in an outsourcing project which resulted in a saving of more than $0.5M during half-year

• Essential contributor in delivery of more than 400 novel genomic targets to pharmaceutical partner, Bayer AG.• Developed computational workflow for identifying potential therapeutic proteins from EST sequences. Performed and delivered automatic annotation of Incyte LifeSeq Gold data to Bayer US Biotech.• Developed system integrating multiple ab initio and homology-based gene-prediction algorithms for identifying novel genomic targets along gene structure curation interface.• Developed protein structure prediction workflow using a combination of FPS and SCOP.