MENTOR: Multiplex Embedding of Networks for Team-Based Omics Research.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Omics technologies generate vast datasets, but methods for analyzing relationships within this data have not kept pace.
• Identifying key biological mechanisms from large-scale omics experiments is crucial for advancing biological discovery.

Purpose of the Study:

• To introduce MENTOR (Multiplex Embedding of Networks for Team-Based Omics Research), a novel network-based algorithm for analyzing omics data.
• To demonstrate MENTOR's capability in both supervised and unsupervised learning contexts for biological interpretation.

Main Methods:

• Developed MENTOR, a network-based algorithm utilizing multiplex embedding.
• Applied MENTOR as a supervised learning method to partition gene sets based on ontological functions.
• Utilized MENTOR as an unsupervised learning method to identify biological functions related to host genetic architecture and microbial abundance in *Populus trichocarpa*.

Main Results:

• MENTOR successfully partitioned gene sets into their respective ontological functions in a supervised learning setting.
• MENTOR identified key biological functions associated with host genetic architecture and microbial abundance in *Populus trichocarpa* using unsupervised learning.
• The open-source nature of MENTOR facilitates distributed interpretation of omics experiments among scientific teams.

Conclusions:

• MENTOR is an effective network-based algorithm for analyzing complex omics data.
• The algorithm advances biological discovery by identifying key mechanisms and facilitating collaborative data interpretation.
• MENTOR provides a valuable tool for both supervised and unsupervised learning in omics research.