sOCP: a framework predicting smORF coding potential based on TIS and in-frame features and effectively applied in the...

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Small open reading frames (smORFs) are increasingly recognized for their roles in critical biological pathways, impacting diseases like diabetes and cancer.
• Accurate in silico prediction of smORFs is essential for analyzing large-scale omics data.

Purpose of the Study:

• To develop and validate a novel computational framework, sOCP, for predicting the coding potential of smORFs.
• To identify novel smORFs in the human genome and create a comprehensive database.

Main Methods:

• Constructed a predictive model for human smORFs using in-frame features and nucleotide bias near the start codon.
• Applied the sOCP model to Rattus norvegicus to assess cross-species performance.
• Scanned the human genome for smORFs, including those with non-canonical start codons, to build a database.

Main Results:

• The sOCP model demonstrated superior prediction accuracy compared to existing methods, with a small feature set preventing overfitting.
• The model showed good performance in both human and Rattus norvegicus.
• Generated a database of approximately one million novel human smORFs with coding potential, including 72,000 located in lncRNA regions.

Conclusions:

• The sOCP framework provides a robust and efficient tool for smORF prediction, enhancing our understanding of their biological significance.
• Identified novel smORFs potentially involved in unique biological processes, such as glucocorticoid catabolism and prokaryotic defense systems.
• The developed database serves as a valuable resource for future research on human smORFs.