Synthetic spike-in standards for RNA-seq experiments.

Area of Science:

• Molecular Biology
• Genomics
• Bioinformatics

Background:

• High-throughput sequencing of complementary DNA (RNA-seq) is a key method for transcriptome profiling.
• Performance variations across different RNA-seq protocols and platforms necessitate standardized evaluation.

Purpose of the Study:

• To develop and utilize synthetic RNA spike-in controls for assessing RNA-seq performance.
• To quantify sensitivity, accuracy, and biases in RNA-seq experiments.
• To establish standard curves for accurate transcript abundance measurement.

Main Methods:

• A pool of 96 synthetic RNAs with diverse lengths and GC content was used as spike-in controls.
• RNA-seq experiments were performed using these controls to measure sensitivity, accuracy, and biases.
• Data from Encyclopedia of DNA Elements (ENCODE) and model organism Encyclopedia of DNA Elements (modENCODE) projects were analyzed.

Main Results:

• Observed linear relationship between read density and RNA input across the detection range.
• Demonstrated high agreement between replicates but greater imprecision than predicted by Poisson errors.
• Identified protocol-dependent biases related to GC content, transcript length, and priming sequences.
• Quantification biases for short transcripts and exons were noted, impacting isoform abundance measurements.

Conclusions:

• Synthetic RNA controls are valuable for evaluating RNA-seq sensitivity and accuracy.
• External RNA controls facilitate comparable analyses across diverse samples, protocols, and platforms.
• Bias correction models can partially mitigate quantification inaccuracies in RNA-seq data.