TSSFinder

TSSFinder predicts transcription start sites (TSS) in eukaryotic genomes, providing ab initio identification of promoter boundaries to support analysis of gene regulation.

Key Features:

• High Accuracy: Provides improved accuracy in predicting TSS signals compared to existing software.
• Speed and Efficiency: Operates with enhanced speed to enable rapid genome analysis.
• Pre-trained Models: Includes pre-trained models for Arabidopsis thaliana, Drosophila melanogaster, Gallus gallus, Homo sapiens, Oryza sativa, and Saccharomyces cerevisiae.
• Customizability: Can be customized for specific organisms using a minimal dataset of 125 DNA sequences with validated TSS signals and their genomic locations.
• Multiple TSS Support: Supports identification of multiple TSSs per gene.

Scientific Applications:

• Genome Annotation: Refines promoter region delineation in genome annotation projects by identifying precise TSS positions.
• Gene Regulation Analysis: Facilitates analysis of transcription regulation by locating TSSs relative to translation start sites.
• Alternative Initiation Studies: Enables study of alternative transcription initiation and multiple promoters per gene by detecting multiple TSSs.

Methodology:

Performs ab initio TSS prediction using machine learning with pre-trained models tailored to specific eukaryotic organisms and allows retraining with a minimal set of 125 DNA sequences with validated TSS positions and genomic coordinates.