Nisha Mulakken

Meta2DB provides a curated metagenomic database with 13,897 samples across 84 studies for predicting human health states using ML models.

The paper introduces an embeddings fusion approach to predict disease state from microbiome features using visual representations of taxonomic data, achieving high classification accuracy across multiple factors.

This study integrates host and sample metadata with microbiome abundance profiles to improve disease prediction models using machine learning.

The paper develops new Centrifuge-compatible nt databases to improve metagenomic classification accuracy by addressing reference database inconsistencies and growth challenges.

The paper assesses how four enterovirus PCR assays capture different species over time using in silico analysis of the NCBI virus database.