Allegro: analyzing expression and sequence in concert to discover regulatory programs

Area of Science:

• Systems biology
• Genomics
• Computational biology

Background:

• Characterizing transcription factors and microRNAs (miRNAs) is crucial for understanding transcriptional regulation.
• Identifying cis-regulatory elements and their associated transcriptional programs remains a challenge in systems biology.

Purpose of the Study:

• To present Allegro, a novel computational method for de-novo discovery of cis-regulatory transcriptional programs.
• To demonstrate Allegro's accuracy and sensitivity compared to existing methods.
• To uncover novel transcriptional modules and regulatory elements across species.

Main Methods:

• Developed Allegro, a log-likelihood-based, non-parametric model for analyzing genome-wide expression data and DNA sequences (promoters, 3' UTRs).
• Applied Allegro to multiple datasets from diverse species, including human stem cells.
• Utilized Allegro for simultaneous analysis of expression data across over 100 conditions.

Main Results:

• Allegro shows higher accuracy and sensitivity than existing techniques for cis-regulatory program discovery.
• Identified a novel motif in murine oocyte gene promoters and new motifs related to fly development.
• Discovered three miRNA families critical for human embryogenesis using stem cell expression profiles.

Conclusions:

• Allegro provides a powerful tool for discovering transcriptional programs and regulatory elements.
• The method facilitates cross-species analysis and identification of key developmental regulators.
• Uncovered novel insights into gene regulation in oocytes, fly development, and human embryogenesis.