TVScript

TVScript quantifies and filters intra-condition read count variation in RNA-Seq datasets to improve the reliability of differential expression analyses.

Key Features:

• Intra-condition variation quantification: Computes transcript-level read count variation across samples within the same condition using RNA-Seq count data.
• Threshold-based filtering: Removes transcripts that exceed user-defined variation thresholds to reduce intra-condition noise.
• Updated count file generation: Produces revised count files with variable transcripts removed for downstream analyses.
• Impact on dispersion estimates: Assesses and improves the distribution of gene-wise dispersion estimates used by most differential expression tools.
• Inter-study and intra-study application: Can be applied at both inter-study and intra-study levels to refine combined or single-study datasets.
• Identification of high-variation DE candidates: Detects genes previously reported as differentially expressed that exhibit high intra-condition variation, flagging their potential unreliability.

Scientific Applications:

• Differential expression analysis: Improves the robustness of differential expression results by reducing intra-condition variability in RNA-Seq count data.
• Transcriptome data refinement: Refines count matrices prior to downstream analyses to enhance consistency of gene expression estimates.
• Comparative studies across populations: Applied to inter-study and intra-study comparisons, including brain samples from dogs, wolves, and foxes (aggressive and tame), to standardize variability across groups.
• Evolutionary and domestication studies: Supports investigation of candidate genes, exemplified by improved support for seven candidate genes implicated in selection for tameness during canine domestication.

Methodology:

Computes per-transcript intra-condition read count variation, filters transcripts exceeding user-defined variation thresholds, outputs updated count files, and evaluates effects on gene-wise dispersion estimates.