Predicting sequence and structural specificities of RNA binding regions recognized by splicing factor SRSF1.

Area of Science:

• Molecular Biology
• Bioinformatics
• Computational Biology

Background:

• RNA-binding proteins (RBPs) are crucial for RNA processing, but their binding specificities are challenging to determine.
• Existing methods for identifying RBP binding sites often neglect RNA structural context.
• Cross-linking immunoprecipitation followed by high-throughput sequencing (CLIP-seq) provides transcriptome-wide binding data.

Purpose of the Study:

• To develop a novel computational model, RNAMotifModeler, for identifying RNA-binding regions.
• To integrate both sequence features and RNA secondary structures for improved motif identification.
• To better understand the sequence specificities that define protein-RNA interactions.

Main Methods:

• Developed RNAMotifModeler, a model-based approach for RNA motif identification.
• Integrated sequence features and RNA secondary structure predictions.
• Applied the model to analyze SRSF1 (SF2/ASF) CLIP-seq data.

Main Results:

• Identified a purine-rich octamer motif 'AGAAGAAG' for SRSF1 binding, preferentially in single-stranded RNA contexts.
• Demonstrated that specific nucleotides and secondary structure elements play complementary roles in SRSF1 binding site recognition.
• Showed higher unpaired probabilities at SRSF1 binding sites compared to controls.

Conclusions:

• Presented a computational model to predict sequence consensus and RNA secondary structures for RBP binding sites.
• The RNAMotifModeler approach shows significant potential for advancing the understanding of RBP binding specificity.
• Successful application to SRSF1 CLIP-seq data validates the model's utility.