Biology subjects

Ranganathan, H.

Publications and source records attributed to Ranganathan, H..

4 recordsLinked to original sources

bioRxiv2026

An Embeddings Fusion Approach Predicts Disease State from Microbiome Features

BackgroundDeep neural networks are a proven technique for working with high dimensional data because of their ability to draw-out meaningful patterns and create vector representations known as "embeddings", which make it easier to work with learning tasks on large inputs as they capture the semantics and variance of the data. Microbial community abundance profiles are well suited for an embeddings approach due to their high dimensionality, and in this work we introduce a novel approach for generating embeddings from visual representations which encode NCBIs taxonomic tree and microbial compositions as images, enabling the creation of embeddings that capture factors such as disease status, type, and geographical location. ResultsWe profiled 13,534 public human metagenomes spanning 85 studies, 24 disease types, 35 countries, and 31,756 microbial species using a profiling pipeline that indexes NCBIs nucleotide database (nt) across all kingdoms of life. Our model achieves an average classification performance of 84% in distinguishing healthy and disease conditions; 87% for disease types, 99% for body sites, and 88% for geographical locations. It also achieves a 97% accuracy when performing multi-label classification of the four factors combined. ConclusionOur work highlights the use of an embeddings approach that can encode multiple features and create efficient contextualization of profiled metagenomes derived from microbiome samples. The models embeddings can be used to cluster existing samples based on multiple conditions and interpretations, and new embeddings can be quickly created for new samples and fitted to existing clusters to characterize them. This has practical applications for unknown, unlabeled microbiome samples.

bioinformatics↗

bioRxiv2025

Beyond Microbial Abundance: Metadata Integration Enhances Disease Prediction in Human Microbiome Studies

Multiple studies have highlighted the human microbiomes potential as a biomarker for diagnosing diseases through its interaction with systems like the gut, immune, liver, and skin via key axes. Advances in sequencing technologies and highperformance computing have enabled the analysis of large-scale metagenomic data, facilitating the use of machine learning to predict disease likelihood from microbiome profiles. However, challenges such as compositionality, high dimensionality, sparsity, and limited sample sizes have hindered the development of actionable models. One strategy to improve these models is by incorporating key metadata from both the host and sample collection/processing protocols. In this paper, we introduce a machine learning-based pipeline for predicting human disease states by integrating host and protocol metadata with microbiome abundance profiles from 68 different studies, processed through a common pipeline. Our findings indicate that metadata can enhance machine learning predictions, particularly at higher taxonomic ranks like Kingdom and Phylum, though this effect diminishes at lower ranks. Our study leverages a large collection of microbiome datasets comprising of 11,208 samples, therefore enhancing the robustness and statistical confidence of our findings. This work is a critical step toward utilizing microbiome and metadata for predicting diseases such as gastrointestinal infections, diabetes, cancer, and neurological disorders.

microbiology↗

bioRxiv2024

Meta2DB: Curated Shotgun Metagenomic Feature Sets and Metadata for Health State Prediction

Meta2DB is a curated metagenomic and metadata database that provides structurally consistent microbiome taxonomy feature count tables for 13,897 samples across 84 studies, 23 disease states, and 34 geographical locations. All samples were uniformly processed using a streamlined metagenomic classification pipeline that employs a reference database indexed to contain all sequences across all kingdoms of life that were present in the NCBI Nucleotide (nt) database retrieved on Jan 04, 2023. This pipeline leverages high-performance computing (HPC) resources at Lawrence Livermore National Laboratory and was used to process 50TB of publicly available raw metagenomic sequence data. Extensive metadata curation was carried out through a combination of manual curation and automated parsing, producing a consistent inter-study metadata table specifically structured to facilitate training of ML models for prediction of human health.

microbiology↗

bioRxiv2022

Model Choice Metrics to Optimize Profile-QSAR Performance

BackgroundPredicting molecular activity against protein targets is difficult because of the paucity of experimental data. Approaches like multitask modeling and collaborative filtering seek to improve model accuracy by leveraging results from multiple targets, but are limited because different compounds are measured with different assays, leading to sparse data matrices. Profile-QSAR (pQSAR) 2.0 addresses this problem by fitting a series of partial least squares models for each target, using as features the predictions from single-task models on the remaining targets. This method has been shown to produce better results than single task and multitask models. However, the factors determining the success of pQSAR 2.0 have as yet not been characterized. In this paper we examine the experimental conditions that lead to better pQSAR models. We limit the amount of data available to the method by retraining with decreasing amounts of data and explore the models ability to generalize to compounds that have never been assayed. Finally, we look at the properties of training data needed to demonstrate pQSAR improvement. ResultsWe apply pQSAR 2.0 on a collection of GPCR and safety targets collected from Drug Target Commons, ExcapeDB, and ChEMBL. We found that pQSAR improved models on 34 of the 149 assays selected. In the other 115 assays, single task random forests offered better performance. There are many factors that contribute to an increase in performance, but the main factor is compound assay coverage. The pQSAR model improves when more compounds are measured in multiple assays. ConclusionIt is necessary to consider the available data before applying pQSAR. Successful pQSAR models require a profile made of correlated targets that share compounds with other assays. This technique is best used when experimental data is available as random forest regressors often do not generalize well enough for virtual drug search applications.

bioinformatics↗