Metagenomics is the study of the structure and function of microbial communities through the application of the whole-genome shotgun (WGS) sequencing method. Providing high-resolution community profiles at species or even strain levels, metagenomics points to a new direction for microbiome research in understanding microbial gene function, microbial-microbial interactions, and host-microbe interactions. My thesis work includes innovation in metagenomic research through the application of ChatGPT in assisting beginning researchers, adopt pre-existed alpha diversity metric for metagenomic data to improve diversity calculation, and the application of metagenomic data in Alzheimer’s disease research.Since the release of ChatGPT in March 2023, the conversation regarding AI in research has promptly been debated. Through the prompted bioinformatic case study, I demonstrate the application of ChatGPT in conducting metagenomic analysis. I constructed and tested a working pipeline aimed at instructing GPT in completing shotgun metagenomic research. The pipeline includes instructions for various essential analytic steps: quality controls, host filtering, read classification, abundance estimation, diversity calculation, and data visualization. The pipeline demonstrated successful completion and reproducible results. Alpha diversity measurement is critical to understanding microbiomes. The widely used Faith’s phylogenetic diversity (PD) metric is agnostic of feature abundance and, therefore, falls short of analyzing metagenomic data. BWPDθ, an abundance weighted variant of Faith’s PD, was implemented in scikit-bio alpha diversity metrics. My analysis shows that BWPDθ does have better performance compared to Faith’s PD, revealing more biological significance, and maintaining their robustness at a lower sampling depth. The progression of Alzheimer’s disease (AD) is known to be associated with alterations in the patient’s gut microbiome. Utilizing metagenomic data from the AlzBiom study, I explored the differential abundance of bacterial pncA genes among healthy and AD participants by age group. The analysis showed that there was no significant difference in pncA abundance between the healthy and AD patients. However, when stratified by age group, within the age group 64 to 69, AD was shown to have significantly lower pncA abundance than the healthy control group. The Pearson's test showed a moderate positive association between age and pncA abundance.