Cufflinks

Cufflinks assembles aligned RNA-Seq reads into a parsimonious set of transcripts and estimates their relative abundances to quantify transcript-level expression and reveal alternative transcription and splicing.

Key Features:

• Read input and parsimonious assembly: Accepts aligned RNA-Seq reads and assembles them into a parsimonious set of transcripts while estimating relative abundances based on read support.
• Transcript discovery and abundance estimation: Performs simultaneous transcript discovery and abundance estimation without requiring prior gene annotations to enable exploration of alternative transcription and splicing events.
• Reference annotation-based transcript assembly: Assembles novel transcripts in the context of existing annotations to address the reference annotation based transcript assembly problem.
• Differential expression analysis (Cuffdiff 2): Uses Cuffdiff 2 to perform transcript-level differential analysis, controlling for variability across replicate libraries and identifying differentially expressed transcripts and genes, as well as changes in splicing and promoter preference.
• Correction of library preparation bias: Applies a likelihood-based approach to account for non-uniform distribution of cDNA fragments within transcripts caused by RNA-Seq library preparation, improving accuracy and replicability across libraries and sequencing technologies.

Scientific Applications:

• Mouse myoblast transcriptome analysis: Applied to over 430 million RNA-Seq reads from a mouse myoblast cell line to detect known and previously unannotated transcripts and to reveal substantial shifts in transcription start sites and splice isoforms during muscle development.
• Lung fibroblast regulatory studies: Used to detect differential expression and splicing changes in lung fibroblasts in response to loss of the HOXA1 transcription factor.
• Transcriptome dynamics and gene regulation: Enables comprehensive characterization of transcriptome dynamics and layers of gene expression regulation from RNA-Seq data.

Methodology:

Accepts aligned RNA-Seq reads, assembles them into a parsimonious transcript set, and estimates transcript abundances from read support; addresses reference annotation-based transcript assembly; performs differential analysis with Cuffdiff 2 controlling for replicate variability; and applies a likelihood-based correction for non-uniform cDNA fragment distribution due to library preparation.