Transcriptome assembly and isoform expression level estimation from biased RNA-Seq reads.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• RNA-Seq (Ribonucleic Acid Sequencing) is a powerful tool for transcriptome analysis but suffers from non-uniform read distribution, introducing biases.
• These biases present significant challenges for accurate gene expression level estimation and transcriptome assembly.
• The impact of these biases on transcriptome assembly has been largely overlooked.

Purpose of the Study:

• To develop a statistical framework for robust transcriptome assembly and isoform expression level estimation from biased RNA-Seq data.
• To model and correct for various RNA-Seq biases, including positional, sequencing, and mappability biases.
• To evaluate the performance of the proposed method on simulated and real-world datasets.

Main Methods:

• Development of a statistical framework utilizing a quasi-multinomial distribution model.
• Implementation of methods to capture and correct for diverse RNA-Seq biases.
• Experimental validation using simulated and real RNA-Seq datasets.

Main Results:

• The proposed method effectively captures various RNA-Seq biases, improving both transcriptome assembly and isoform expression level estimation.
• Experimental results demonstrate high sensitivity and precision in transcriptome assembly.
• Estimated expression levels show high concordance with quantitative reverse transcription-polymerase chain reaction (RT-qPCR) data.

Conclusions:

• The developed statistical framework provides a robust solution for analyzing biased RNA-Seq data.
• The method enhances the accuracy of transcriptome assembly and gene expression quantification.
• This work offers a valuable tool for researchers in transcriptomics and related fields.