Investigating cis-regulatory elements and gene expression in multiple tomato varieties using interpretable deep...

Area of Science:

• Genomics
• Computational Biology
• Plant Science

Background:

• Gene expression is regulated by cis-regulatory elements (CREs), but their precise influence is complex.
• Understanding non-coding regulatory elements is crucial for deciphering gene expression patterns.

Purpose of the Study:

• To develop an interpretable deep learning model (L-CRE) for predicting gene expression levels based on CREs.
• To identify critical genomic regions and regulatory mechanisms influencing gene expression in tomatoes.

Main Methods:

• Refined existing models and developed the L-CRE deep learning framework.
• Analyzed gene flanking regions from four tomato varieties.
• Performed interpretability analysis to evaluate genomic region contributions to predictions.

Main Results:

• L-CRE achieved a peak prediction accuracy of 86.9% for high and low gene expression levels.
• Identified critical genomic regions significantly impacting gene expression predictions.
• Discovered that these regions often contain transcription factor binding sites.
• Validated identified regulatory elements through experimental data.

Conclusions:

• The L-CRE model provides a robust method for predicting gene expression and understanding regulatory mechanisms.
• Identified key genomic regions and transcription factor binding sites offer insights into tomato gene regulation.
• This research advances functional genomics and provides tools for crop genetic improvement.