Wenxing Hu

● Conducted in-depth analyses of 10+ large scale NGS single cell RNA-seq and spatial transcriptomics datasets to identify therapeutic targets. ● Developed platforms for centralized omics data storage, processing pipeline, and visualization tools. ● Developed faster (100X) and scalable (20 million cells) analytical package for scRNA-seq data. Tools used: GPU computing, CuPy, CUDA kernel. https://github.com/interactivereport/scaleSC ● Deployed scGPT, LLM models, chatbots for internal data applications. ● Disease diagnosis using vision Transformers and retina images. Image processing, image segmentation, image biomarker detection. https://github.com/interactivereport/FundusNet ● Designed small molecules using LLMs and Reinforcement Learning techniques. ● Collaborated with cross-functional teams and supervised interns and co-ops in target discovery and validation through multiomics and computational biology approaches.

Implemented contrastive learning method to mental disorder studies Developed deep learning models for brain imaging analysis

* Conducted data cleaning of lung cancer clinical data and gene expression data. * Implemented a multi-modality deep learning framework to integrate gene expression data and CT images to predict cancer clinical outcomes. * Applied U-net for lesion segmentation. * Used Transfer learning and Auto-encoder network to achieve a better prognosis.