Built a custom Transformer model achieving 99.67% accuracy in character frequency classification and extended it to a full language model with a low perplexity of 6.13 for next-character prediction, using Transformer-based sequence modeling for accurate and efficient language prediction.

Developed a pipeline using unsupervised machine learning to identify six astrocyte subtypes through spatial clustering of spatial transcriptomic data from the Allen Mouse Brain Atlas, discovered 104 astrocyte-specific genes through differential expression analysis with Bonferroni correction across 10+ million cells.