The Flux Simulator

The Flux Simulator simulates RNA-Seq experiments in silico to model production of sequencing reads from annotated transcripts on a reference genome and to analyze technical biases across reverse transcription, fragmentation, adapter ligation, PCR amplification, gel segregation, and sequencing.

Key Features:

• Simulation of Technical Components: Deconstructs RNA-Seq experiments into reverse transcription, fragmentation, adapter ligation, PCR amplification, gel segregation, and sequencing and models each step's effect on read abundance and distribution.
• Universally Applicable Models: Provides parameterizable models for each technical component that use empirical attributes from specific experimental protocols.
• Reproduction of Experimental Data: Generates read distributions by simulating combinations of technical steps that closely match distributions observed in actual experiments.
• Insight into Hidden Precursors: Offers insights into hidden precursors that affect the final configuration of reads along gene bodies to explain contradictory read distribution observations.
• Identification of Systematic Biases: Identifies sources of systematic bias, including biases arising from RNA hydrolysis fragmentation.

Scientific Applications:

• Bias Analysis: Investigate and mitigate biases introduced at stages such as reverse transcription, fragmentation (including RNA hydrolysis), adapter ligation, PCR amplification, gel segregation, and sequencing.
• Protocol Optimization: Simulate protocol configurations to identify parameter adjustments that reduce bias or improve experimental outcomes.
• Data Interpretation: Reveal hidden factors and precursor effects along gene bodies to explain complex read distribution patterns in RNA-Seq data.

Methodology:

The simulator deconstructs RNA-Seq experiments into technical components, applies parameterizable models using empirical protocol attributes, simulates combinations of the modeled steps to generate sequencing read distributions, and compares simulated distributions to experimental data to identify systematic biases such as those from RNA hydrolysis.