Long-range regulation is a major driving force in maintaining genome integrity

Area of Science:

• Genomics
• Evolutionary Biology
• Bioinformatics

Background:

• Comparative genomics has expanded due to new vertebrate genome sequences and improved alignment algorithms.
• Large-scale genome rearrangements alter gene order, with some chromosomal regions acting as hotspots for breakpoints.
• The reasons for the uneven distribution of breakpoints during vertebrate evolution remain unclear.

Purpose of the Study:

• To develop a machine learning method for distinguishing between genomic regions prone to deleterious breakpoints (refractory) and those with neutral breakpoints (susceptible).
• To identify features characterizing these distinct genomic regions.

Main Methods:

• A machine learning predictor was trained using breakpoint data from the human lineage since amniote divergence.
• Analysis of features distinguishing breakpoint-refractory and breakpoint-susceptible regions.

Main Results:

• Breakpoint-refractory regions are significantly enriched for conserved non-coding elements and developmental genes.
• Breakpoint-susceptible regions show enrichment for housekeeping genes and simpler transcriptional regulation.
• The predictor successfully differentiates between refractory and susceptible regions based on these features.

Conclusions:

• Long-range transcriptional regulation is a key factor influencing the fixation of chromosome breaks.
• High fitness costs associated with disrupting spatial associations between regulatory elements and target genes prevent rearrangements in many regions.
• Only a specific, identifiable portion of the genome is susceptible to large-scale rearrangements.