BiomiX 3.0: A user-friendly platform for democratized multi-omics integration with graph-based learning.
Background Multi-omics integration has emerged as a powerful strategy to decode the molecular complexity of biological systems. However, the diversity of available methods, each designed with distinct assumptions, objectives, and computational requirements, makes method selection, usage and interpretation challenging for nonexpert users. Here we present BiomiX 3.0, an updated version of the…
Multi-omics integration is a powerful approach for understanding the complexity of biological systems, but the multitude of methods available can be overwhelming for non-expert users. The BiomiX platform, now in its updated version 3.0, addresses this challenge by offering a user-friendly graphical user interface (GUI) for integrating various multi-omics datasets.
BiomiX 3.0 extends its capabilities with four additional integration methods: Similarity Network Fusion (SNF), NEighborhood-based Multi-Omics clustering (NEMO), DIABLO for biomarker discovery, and PRAMIGO, a novel supervised heterogeneous graph transformer. The platform was benchmarked on two independent multi-omics datasets: one consisting of Chronic Lymphocytic Leukemia (CLL) patients and another of pulmonary tuberculosis (PTB) patients versus healthy controls.
Comparing supervised and unsupervised methods, the results showed that supervised methods (DIABLO and PRAMIGO) consistently outperformed unsupervised methods (SNF and NEMO) in condition-specific discrimination. Unsupervised methods, however, revealed alternative patient stratifications based on independent sources of biological variance, while MOFA performed semi-supervisedly, capturing latent factors that explain both disease-associated and orthogonal sources of variance.
By analyzing the top-ranked features prioritized by each method, the study identified 154 shared features across both cohorts, spanning B-cell receptor biology, innate immune signaling, RAS/MAPK activation, and epigenetic regulation in CLL, and a convergent interferon/innate immune signature in PTB. Additionally, method-specific features were discovered, such as vesicle trafficking in DIABLO, immune checkpoints in MOFA, ncRNA regulation in PRAMIGO, acylcarnitine metabolic reprogramming in DIABLO, restoration of lysosomal trafficking in MOFA, and {gamma}{delta} T-cell and immunoglobulin repertoire diversity in PRAMIGO.
Most notably, PRAMIGO stands out for its ability to model cross-modal molecular interactions through heterogeneous graphs, uncovering complex epigenetic co-regulation programs in CLL and multi-omics inflammatory modules in PTB that are difficult to identify using conventional integration approaches. Ultimately, BiomiX 3.0 democratizes access to advanced multi-omics analysis, enabling researchers across disciplines to interrogate biological variation in their data without requiring specialized bioinformatics expertise.
Written by urgent.news from bioRxiv's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
