Advised by Dr. Lulu Shang

• Developing MANGO, a multiscale permutation framework for detecting spatial relationships among multiple cell types.
• Validated the framework across 2D and 3D spatial omics datasets from cancer and neural tissues.
• Evaluated clinical relevance by testing associations between spatial cell-type patterns, cancer stage, and patient survival.

Advised by Dr. Sizun Jiang

• Developed a question-generation and benchmarking framework to compare ProbeGPT's scientific question-answering performance with general-purpose language models.
• Built automated workflows to retrieve manuscripts, protocols, manuals, and webpages, expanding the RAG knowledge base from 132 to 1,900 documents.
• Validated the INSIDE spatial multi-omics platform across H&E imagery, CODEX protein markers, and clinical metadata.

Collaboration with Dr. Matthew N. Rasband

• Performed Palm2 alternative splicing analysis using GSE185862 single-cell RNA-seq data.
• Built an RNA-seq preprocessing and splice-junction mapping pipeline to quantify cell-type-specific exon splicing patterns.

Advised by Dr. Anindita Basu

• Selected for a competitive NITMB summer fellowship to investigate abnormal hemodynamics after arteriovenous fistula creation.
• Developed PyTorch convolutional neural networks to classify blood-flow microscopy images as laminar or non-laminar.
• Compared model performance and created visual explanations for experimental and medical collaborators.

Advised by Dr. Evan Siemann

• Investigated how geographic variation in soil microbial communities affects Chinese tallow tree invasion intensity.
• Quantified arbuscular mycorrhizal fungi colonization frequency in plant-root samples through microscopy.
• Analyzed relationships between fungal colonization and variation in soil microbial communities.