Exploiting hidden information interleaved in the redundancy of the genetic code without prior knowledge

Area of Science:

• Computational Biology
• Genomics
• Molecular Biology

Background:

• Codon usage influences gene expression regulation across all stages.
• Analyzing gene expression requires high-quality, large-scale expression data for model inference.
• Lack of such data limits the ability to infer and validate gene expression models.

Purpose of the Study:

• To introduce Chimera, an unsupervised computational approach for analyzing gene expression encoded in open reading frames (ORFs).
• To develop Chimera Average Repetitive Substring (ChimeraARS) for estimating ORF adaptability to host gene expression machinery.
• To develop ChimeraMap for engineering codons to improve adaptation to new host gene expression machinery.

Main Methods:

• Chimera exploits hidden high-dimensional information within an organism's genome.
• ChimeraARS computes the tendency of ORFs to include long, repetitive substrings within coding sequences.
• ChimeraMap engineers codons to incorporate frequent host coding sequence substrings.

Main Results:

• Chimera is computationally efficient and requires only genome sequence data.
• ChimeraARS demonstrated significant correlations (up to 0.67) with heterologous gene expression in *Escherichia coli*.
• Chimera identified signals undetectable by conventional methods like the Codon Adaptation Index (CAI).

Conclusions:

• Chimera offers a novel approach for analyzing and engineering gene expression, especially when expression data is limited.
• The Chimera approach, including ChimeraARS and ChimeraMap, enhances the understanding and manipulation of gene expression.
• Chimera provides a powerful tool for analyzing both endogenous and heterologous genes, outperforming existing methods in certain contexts.